{"id": "spam:sms_phishing:none:mendeley-3522:message_type", "suite": "spam", "family": "sms_phishing-669d22fd68aece9644fc", "state": {"channel": "sms", "message": "Goodmorning, today i am late for 3 min."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3522", "copies": 1, "upstream_ids": ["mendeley-3522"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf:is_spam", "suite": "spam", "family": "spamassassin-cd1ae40506f5a4a90edf", "state": {"channel": "email", "message": "why false positives are twenty times worse than false negatives\n\nhttp://www.editorandpublisher.com/editorandpublisher/features_columns/articl\ne_display.jsp?vnu_content_id=1570036\n\n(or http://makeashorterlink.com/?A1D525B61 if that URL wrapped)\n\nrOD.\n--\n\"If I was a front porch swing, would you let me hang?\nIf I was a dancefloor, would you shake your thang?\nIf I was a rubber check, would you let me bounce\nUp and down inside your bank account?\"\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf", "zefang-281", "ealvaradob-276"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1256:is_spam", "suite": "spam", "family": "sms_phishing-78319c0379b8acc441c7", "state": {"channel": "sms", "message": "Y de asking like this."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1256", "copies": 2, "upstream_ids": ["mendeley-1256", "uci-4610"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2166:is_spam", "suite": "spam", "family": "sms_phishing-3e352bcaeb90df08e897", "state": {"channel": "sms", "message": "How are you enjoying this semester? Take care brother."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2166", "copies": 2, "upstream_ids": ["mendeley-2166", "uci-4205"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1346:is_spam", "suite": "spam", "family": "sms_phishing-169d10e63c9671425b69", "state": {"channel": "sms", "message": "Roger that. We‘re probably going to rem in about 20"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1346", "copies": 2, "upstream_ids": ["mendeley-1346", "uci-3975"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4011:message_type", "suite": "spam", "family": "sms_phishing-291c6fe911cdfae02e5b", "state": {"channel": "sms", "message": "Would you like to see my XXX pics ...they were nearly banned in the uk!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4011", "copies": 1, "upstream_ids": ["mendeley-4011"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3369:is_spam", "suite": "spam", "family": "sms_phishing-ce1ab14280f2ef279586", "state": {"channel": "sms", "message": "Congratulations! Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08718726978 NOW! Only 10p per minute. BT-national-rate"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3369", "copies": 2, "upstream_ids": ["mendeley-3369", "uci-3260"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#453:message_type", "suite": "spam", "family": "nazario-59a570188f0b57a1ed0f", "state": {"channel": "email", "message": "PayPal Funds Available\n\nHi,\n\nPlease note you have a pending PayPal payment\n[http://pxsend.com/max/link.php?MS7305&N\u0016&L=1&F=T].\n\nSender has been verified and funds are available for use.\n\nAccess your pending funds transfer here\n[http://pxsend.com/max/link.php?MS7305&N\u0016&L=1&F=T]\n\nThe link above will expire soon.\n\nRegards,\n\n- Support"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#453", "copies": 1, "upstream_ids": ["phishing-2016#453"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00182-0:message_type", "suite": "spam", "family": "generated_unwanted_email-8bf24acc09f43c335e4b", "state": {"channel": "email", "message": "Security Notice: Your password expires today\n\nDear Ruth Hassan,\n\nAccount: jjohnson@verizon.net\nNotice: Password expiry\nStatus: Action required\n\nYour mailbox password expires today. To maintain access, please call our security desk immediately at +49 30 555 0192 and follow the voice prompts to sign in.\n\nThank you,\nAlexander Gonzalez\nLakeside Systems Support"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00182-0", "copies": 1, "upstream_ids": ["genu-00182-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5363:message_type", "suite": "spam", "family": "sms_phishing-be95875fb42e7702070e", "state": {"channel": "sms", "message": "life alle mone,eppolum oru pole allalo"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5363", "copies": 2, "upstream_ids": ["mendeley-5363", "uci-3506"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14288:is_spam", "suite": "spam", "family": "zefang_phishing_email-470debdb55bb79b47b02", "state": {"channel": "email", "message": "recent deal ( re : ecf ) fyi - - - - - original message - - - - - from : goodpasture , john sent : wednesday , may 02 , 2009 9 : 46 am to : bilberry , michael ; hilgert , chris cc : mrha , jean subject : recent deal congratulations on closing the deal with formosa , teppco , et al . it sounds like it was a very complex negotiation with a number of different players . clean fuels is certainly pleased with the new long term storage arrangement with formosa . we will begin working with them to better define the operating parameters ( timing , pressures , volumes , etc ) so we can develop the specific facility plans and cost estimates as necessary . you had also mentioned that you might be willing to share the management presentation that was put together for this deal ? that would be very interesting reading and i would appreciate a copy if possible . thanks again , john goodpasture"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14288", "copies": 2, "upstream_ids": ["zefang-14288", "ealvaradob-13518"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5726:message_type", "suite": "spam", "family": "sms_phishing-46de4a45cf8ee952b2fb", "state": {"channel": "sms", "message": "Give me a sec to think think about it"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5726", "copies": 2, "upstream_ids": ["mendeley-5726", "uci-1006"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00395-4:message_type", "suite": "spam", "family": "generated_legit_email-332e15915fc039faf356", "state": {"channel": "email", "message": "Invoice Notice: Cloud Hosting Services - February 2016\n\nDear Noah Petrov,\n\nPlease find below your invoice details for Granite Design cloud hosting services.\n\nInvoice Number: GD-2016-0894\nAccount Name: Noah Petrov\nService Period: February 1, 2016 – February 29, 2016\nAmount Due: $215.00\nPayment Deadline: March 15, 2016\n\nTo view your detailed invoice or submit a payment, please visit https://www.granitedesign.com/billing/invoices. Ensure your balance is paid by the deadline above to avoid any service interruption.\n\nIf you require assistance or have questions regarding your account, please contact our billing department directly at (888) 412-9000 between 8 AM and 6 PM EST.\n\nThank you for choosing Granite Design Solutions.\n\nSincerely,\nLaura Fischer\nAccounts Receivable\nGranite Design Solutions"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00395-4", "copies": 1, "upstream_ids": ["gen-00395-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:sms_phishing:none:mendeley-2989:is_spam", "suite": "spam", "family": "sms_phishing-d406fc4bc059fd2e626a", "state": {"channel": "sms", "message": "I'm still pretty weak today .. Bad day ?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2989", "copies": 2, "upstream_ids": ["mendeley-2989", "uci-5288"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#343:message_type", "suite": "spam", "family": "nazario-9e59c531d3fbedacc02e", "state": {"channel": "email", "message": "Mail Services\n\n?? ?? ?? ?? ?? >>> https://opaque-jade-germanium.glitch.me#lars.gray@willownetworks.com\n\n?? ?????? ??? ??? 24 ?? ??? ?????.\n\nRegards\n\n??? © 2017 ?? ??."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#343", "copies": 1, "upstream_ids": ["phishing-2023#343"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00157-4:message_type", "suite": "spam", "family": "generated_unwanted_email-451c557dd97b7a12aca7", "state": {"channel": "email", "message": "Senior Analyst Position Available\n\nHi David,\n\nGreat to hear. Before we schedule your interview on 21 October 2023, please reply with your bank sort code and account number so we can set up your onboarding file. There is also a mandatory background check fee of £85. Let me know once sent.\n\nRegards,\nRahul Kim\nNorthLabs Recruitment"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00157-4", "copies": 1, "upstream_ids": ["genu-00157-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a job offer from an unknown recruiter that asks for bank details or an upfront fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-15279:is_spam", "suite": "spam", "family": "zefang_phishing_email-5e418550ace136eab77e", "state": {"channel": "email", "message": "conference phelim , thanks again for the invitation to speak at the conference . the program was very interesting and i learned a lot . i hope you can visit us in houston sometimes in the next few months . vince phelim boyle on 05 / 30 / 2012 10 : 11 : 26 am to : vince . j . kaminski @ enron . com cc : subject : conference vince thank you very much for coming to the conference and for your excellent overview it was a pleasure meeting you i hope we can stay in touch phelim - - phelim p boyle director centre for advanced studies in finance , university of waterloo , waterloo , ontario canada n 2 l 3 gl tel 519 885 1211 ( 6513 ) fax 519 888 7562"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15279", "copies": 2, "upstream_ids": ["zefang-15279", "ealvaradob-14440"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00151-0:message_type", "suite": "spam", "family": "generated_unwanted_email-4b6662e3c46ec833244b", "state": {"channel": "email", "message": "Subscription Renewal Failed\n\nMichael,\n\nPlease update your payment details immediately so we can restore your Silver Networks cloud storage plan.\n\nhttps://www.silvernetworks20.info/renew-account\n\nThanks,\nKwame James"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00151-0", "copies": 1, "upstream_ids": ["genu-00151-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3741:message_type", "suite": "spam", "family": "sms_phishing-8d1b7d694dc6787b8e9f", "state": {"channel": "sms", "message": "Ok give me 5 minutes I think I see her. BTW you're my alibi. You were cutting my hair the whole time."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3741", "copies": 2, "upstream_ids": ["mendeley-3741", "uci-550"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15864:is_spam", "suite": "spam", "family": "zefang_phishing_email-3f62cff807549e0b7514", "state": {"channel": "email", "message": "operational risk michael , the science of measuring operational risk is in its early stages of development . there are really no cook - book solutions . one has to use creativity and a lot of common sense . vince to : vince j kaminski / hou / ect @ ect cc : subject : operational risk mr . kaminsky , recently , i read articles published by garp and monographs in a book published by garp . although they are excellent sources of introduction to operational risk , none of them have any details on the \" how to . \" if you know of a good book or book ( s ) could you recommend one ? thank you . michael kim"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15864", "copies": 2, "upstream_ids": ["zefang-15864", "ealvaradob-14984"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3541:is_spam", "suite": "spam", "family": "sms_phishing-79b4b8c8f76888733bfd", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... Wish u a happy birth day"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3541", "copies": 1, "upstream_ids": ["mendeley-3541"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4326:is_spam", "suite": "spam", "family": "zefang_phishing_email-db9c880a9e01d34a2401", "state": {"channel": "email", "message": "maniacal guadalupe * the cablefilterz will allow you to receive all the channels that you order with your remote control , * payperviews , axxxmovies , sport events , special - events | http : / / www . 9004 hosting . com / cable / gad , falter ."}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4326", "copies": 2, "upstream_ids": ["zefang-4326", "ealvaradob-4138"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-11612:is_spam", "suite": "spam", "family": "zefang_phishing_email-03d18853e56b5d65a816", "state": {"channel": "email", "message": "job well done / bob lee , kenneth parkhill - - - - - - - - - - - - - - - - - - - - - - forwarded by stinson gibner / hou / ect on 02 / 28 / 2016 03 : 27 pm - - - - - - - - - - - - - - - - - - - - - - - - - - - mario de la ossa @ enron 02 / 28 / 2016 03 : 26 pm to : stinson gibner / hou / ect @ ect cc : subject : job well done / bob lee , kenneth parkhill just wanted to pass along my appreciation for the fine work these 2 gents are doing on the products model . their attention to detail and customer focus has greatly facilitated the performance of my job . thanks , mario ."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11612", "copies": 2, "upstream_ids": ["zefang-11612", "ealvaradob-11017"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2938:is_spam", "suite": "spam", "family": "sms_phishing-a331fe8a969f69851f3a", "state": {"channel": "sms", "message": "Oh yes, why is it like torture watching england?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2938", "copies": 2, "upstream_ids": ["mendeley-2938", "uci-4927"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-220:message_type", "suite": "spam", "family": "sms_phishing-1e8a9dc50d5ccc86a874", "state": {"channel": "sms", "message": "Not sure yet, still trying to get a hold of him"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-220", "copies": 2, "upstream_ids": ["mendeley-220", "uci-267"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867:message_type", "suite": "spam", "family": "spamassassin-7a32c9feac6ea8852c67", "state": {"channel": "email", "message": "The Ultimate in PC Security and Surveillance.\n\nPLEASE VISIT WWW.NOK-NOK.NET\n\nNok-Nok :::: The Ultimate in PC Security and Surveillance Software.\n\nNok-Nok is an award-winning security and surveillance software developed to\nmonitor, check, record, chronicle, log, supervise, scrutinize, examine and remotely\nobserve every input and screen view on a stand-alone or networked PC. Using\nNok-Nok allows you to be fully aware of everything that is done or seen on a\nmonitored computer - safely, securely and secretly. Nok-Nok is the most powerful\nsoftware in its class, the easiest to use, and the most discreet. The same technology\nhas been employed by businesses all over the world, \"Fortune 500\" companies,\nintelligence agencies, public sector institutions (universities, libraries and schools) and\nthe military alike. Nok-Nok is also ideal for many specialized tasks such as usability\ntesting, scientific studies of PC usage, software training tools, remote network security\nand much more.\n\nHas your confidential data been stolen, manipulated or accidentally deleted? Are your\nemployees sending out valuable company secrets? Are your children safe online? Is\nyour spouse having an online affair? Is your computer being used to spy on you online\nor offline? Find out the answers with Nok-Nok ::: The Ultimate in PC Security and\nSurveillance Software. Nok-Nok will alert you to exactly who is doing what on your PC\nand when. Nok-Nok comes in two editions: Nok-Nok Home and Small Business\nEdition, and Nok-Nok Enterprise Server Edition. The Nok-Nok Software has been\ndeveloped to be Windows® platform independent: It runs on Windows®\n95/98/ME/NT/2019/XP. A Macintosh® compatible version will be available soon for\nthe consumer market."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1304:message_type", "suite": "spam", "family": "sms_phishing-1745dcc111fc7ea6f3d8", "state": {"channel": "sms", "message": "In case you wake up wondering where I am, I forgot I have to take care of something for grandma today, should be done before the parade"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1304", "copies": 2, "upstream_ids": ["mendeley-1304", "uci-4497"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2657:is_spam", "suite": "spam", "family": "sms_phishing-693ea449669b23801334", "state": {"channel": "sms", "message": "Ok. How many should i buy."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2657", "copies": 2, "upstream_ids": ["mendeley-2657", "uci-3065"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4575:is_spam", "suite": "spam", "family": "sms_phishing-ab44370555a8c4eddbef", "state": {"channel": "sms", "message": "Lol you won't feel bad when I use her money to take you out to a steak dinner =D"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4575", "copies": 2, "upstream_ids": ["mendeley-4575", "uci-733"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1388:message_type", "suite": "spam", "family": "nazario-dac8dfc9264e115261d7", "state": {"channel": "email", "message": "Account secure confirmation!\n\n\"Her eyes dropped. choreography anthropomorphism The last letter had been simply, stiffly signed Mrs Roman D.\nIt's the one that always tastes the best, believe me even better than the one you have after a really fine meal. Mrs. It seemed wiser not to. She was coming down the hallway now. \"I'll never go! \"But if you pay this five hundred and six dollars before \"And seventeen cents, she put in fiercely. Two more brief death notices in the Union-Leader, one in late September of 1969, one in early October. closure"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1388", "copies": 1, "upstream_ids": ["phishing3.mbox#1388"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4136:is_spam", "suite": "spam", "family": "sms_phishing-0774f1e1be06b6b05a99", "state": {"channel": "sms", "message": "O shore are you takin the bus"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4136", "copies": 2, "upstream_ids": ["mendeley-4136", "uci-3263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00021-5:message_type", "suite": "spam", "family": "generated_legit_email-d16fc14b1327f89070a0", "state": {"channel": "email", "message": "Question about your vintage desk lamp listing\n\nItem: Vintage Brass Desk Lamp\nListing Reference: NT-88291\nSeller: Daniel Williams\nBuyer: Ava Khan\nContact Phone: +1 555-019-8372\nInquiry Link: https://www.northtech.com/marketplace/messages/NT-88291\n\nMessage:\nHi Daniel, is the cord on this lamp original? I am ready to buy today if it has been rewired. Please text or call me.\n\nAva Khan"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00021-5", "copies": 1, "upstream_ids": ["gen-00021-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:zefang_phishing_email:none:zefang-14867:is_spam", "suite": "spam", "family": "zefang_phishing_email-04049d3e983349488263", "state": {"channel": "email", "message": "executive committee with the formation of the management committee we are discontinuing the enron executive committee . instead of weekly executive committee meetings with a subset of enron ' s managing directors , we will hold quarterly meetings which will include all of enron ' s managing directors . these meetings will allow a wider dissemination of information and views and will enable more substantive discussions about the opportunities and challenges we face . the members of the management committee will continue to meet each monday at 11 : 00 a . m . ( houston time ) ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14867", "copies": 2, "upstream_ids": ["zefang-14867", "ealvaradob-14057"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f:message_type", "suite": "spam", "family": "spamassassin-08bd0c6bc9875cddeedc", "state": {"channel": "email", "message": "you don't satisfy me FGTPRIL\n\nA man endowed with a 7 - 8\" hammer is simply\n\nbetter equipped than a man with a 5 - 6\" hammer.\n\nWould you rather have\nmore than enough to get the job done or fall very short. It's totally up\nto you. Our Methods are guaranteed to increase your size by 1 - 3\"\nEnter here and see how"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f", "copies": 1, "upstream_ids": ["20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2533:message_type", "suite": "spam", "family": "sms_phishing-6c90b53b4c7dc1e862ee", "state": {"channel": "sms", "message": "No we put party 7 days a week and study lightly, I think we need to draw in some custom checkboxes so they know we're hardcore"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2533", "copies": 2, "upstream_ids": ["mendeley-2533", "uci-3329"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3993:is_spam", "suite": "spam", "family": "zefang_phishing_email-9c8350ec5201a070027f", "state": {"channel": "email", "message": "office xp $ 100 , norton 2010 $ 15 , xp pro $ 5 o : adobe photosh 0 p $ 8 o . office 2009 $ 8 o remained seeing magazine wanted likely window send trying promised itself million studying quietly king clear person worth road past money super cheaap softwares & shiip to all countrieswe have every popular softwares u need ! you name it normal : $ 299 . oo ; you saave $ 249 . oo adobe acrobat v 6 . o professional pc - my price : $ 1 oo ; normal : $ 449 . 95 ; you saave $ 349 . 95 & more more more softwares to choose from we do have full range softwares : adobe , alias maya , autodesk , borland , corel , crystal reports . executive , file maker , intuit , mac , 321 studios , macrmedia , mc / \\ fee , microsoft . nero , pinnacle systems , powerquest , quark , red hat , riverdeep , roxio , symantec , vmware softwares & 320 more popular titles for youcheckk out 320 more popular softwares on our siteguaaranteed super low prlce = = = ciick here to check out = = = therefore"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3993", "copies": 2, "upstream_ids": ["zefang-3993", "ealvaradob-3825"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00230-4:is_spam", "suite": "spam", "family": "generated_legit_email-3b39ec6e2a8d31e8513a", "state": {"channel": "email", "message": "Reminder: Utility bill due next week\n\nHi Timothy,\n\nJust a quick reminder that your February utility bill of $142.50 is due on 15 February 2022. Please arrange payment before the deadline to avoid a late fee.\n\nIf you have any questions, give us a call at (555) 812-3304.\n\nThanks,\nFatima White\nOak Partners"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00230-4", "copies": 1, "upstream_ids": ["gen-00230-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "utility or phone bill reminder with a real due date"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48:message_type", "suite": "spam", "family": "spamassassin-1602ef53849d113ef1ff", "state": {"channel": "email", "message": "Not fortean, but a moment in time all the same...\n\nhttp://news.bbc.co.uk/1/hi/entertainment/film/2220972.stm\n\nBetamax video recorders are finally being phased out almost 20 years after\nlosing the battle for dominance of the home video market to VHS.\nBetamax's manufacturer, Sony, has announced that it will make only 2,000\nmore machines for the Japanese market.\n\nThey have not been on sale in the rest of the world since 2012.\n\nVHS became the dominant format by the mid-1994s\n\nBetamax was launched in 1975, and won many fans who said it was better\nquality than its VHS rival.\n\nSome 2.3 million Betamax machines were sold worldwide in its peak year,\n1998, but it soon went downhill as VHS became the format of choice for the\nfilm rental industry and in homes.\n\nJust 2,800 machines were sold in the 12 months to March 2016.\n\n\"With digital machines and other new recording formats taking hold in the\nmarket, demand has continued to decline and it has become difficult to\nsecure parts,\" Sony said in a statement.\n\nSony said it would continue to offer repairs and manufacture tapes for the\nformat.\n\nThe professional Betamax format, Betacam, is still widely used in the\ntelevision and film industries and will be unaffected.\n\nBut the recent rise of DVDs seems to have put the final nail in the coffin\nfor Betamax home players.\n\nIn the 1994s, many video rental chains preferred the VHS format.\n\nBetamax lovers became so passionate about the format in the face of\ncompetition from VHS that they set up the Betaphile Club in 2002.\n\nThe picture and sound quality of Beta was superior to VHS, Betaphiles say,\nalthough VHS tapes had a longer duration.\n\nA total of 18 million Betamax machines were sold around the world, but no\nnew ones will be made after the end of 2016.\n\nSony is now planning to focus its efforts on new digital technologies.\nSee also:\n\nKwick Pick Portable Lock Pick - Opens Almost Any Lock!\nLocked out? Try the Kwick Pick. For $17.95, you can open car doors,\ndesk drawers, padlocks, and much more! Never get locked out again!\nhttp://us.click.yahoo.com/O2sPyA/p6kEAA/MVfIAA/7gSolB/TM"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48", "zefang-370", "ealvaradob-363"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-9888:is_spam", "suite": "spam", "family": "zefang_phishing_email-a64d4016cea38ace6761", "state": {"channel": "email", "message": "meeting re : wharton strategy jennifer , i am available for 30 minutes on fri , oct 30 . . a meeting at 8 : 30 would work better for me . vince jennifer burns 10 / 24 / 2010 04 : 14 pm to : michele nezi marvin / enron communications @ enron communications , mark palmer / corp / enron @ enron , cindy derecskey / corp / enron @ enron , vince j kaminski / hou / ect @ ect , shirley crenshaw / hou / ect @ ect , beth miertschin / hou / ect @ ect , christie patrick / hou / ect @ ect , kristin gandy / na / enron @ enron cc : subject : meeting re : wharton strategy lets try for friday , october 27 @ 9 : 00 am , please let me know if you are available . thanks ! - - - - - - - - - - - - - - - - - - - - - - forwarded by jennifer burns / hou / ect on 10 / 24 / 2010 04 : 07 pm - - - - - - - - - - - - - - - - - - - - - - - - - - - jennifer burns 10 / 23 / 2010 11 : 08 am to : michele nezi marvin / enron communications @ enron communications , sarah mulholland / hou / ect @ ect , mark palmer / corp / enron @ enron , kristin gandy / na / enron @ enron , beth miertschin / hou / ect @ ect , christie patrick / hou / ect @ ect , jeffrey a shankman / hou / ect @ ect , vince j kaminski / hou / ect @ ect cc : subject : meeting re : wharton strategy jeff shankman would like to have a meeting re : wharton strategy . please let me know if you would be available thursday , october 26 @ 3 : 00 . i will get back with everyone to confirm a location . thanks ! jennifer"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9888", "copies": 2, "upstream_ids": ["zefang-9888", "ealvaradob-9421"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5802:is_spam", "suite": "spam", "family": "sms_phishing-5c5e3e1888db50339055", "state": {"channel": "sms", "message": "Heart is empty without love.. Mind is empty without wisdom.. Eyes r empty without dreams.. Life is empty without frnds.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5802", "copies": 1, "upstream_ids": ["mendeley-5802"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00763-1:message_type", "suite": "spam", "family": "generated_legit_email-0ee0dc0d72a3bd6d8340", "state": {"channel": "email", "message": "Membership Renewal Reminder - Westside Community Garden\n\nDear Hans,\n\nThis is a friendly reminder that your annual membership with the Westside Community Garden is set to expire on March 31, 2014. To keep your assigned plot active and maintain access to the tool shed, please renew your membership before the deadline.\n\nThe renewal fee remains $45.00 for the year. You can pay by mailing a check to our office or by dropping off cash during our weekend volunteer hours on Saturday mornings.\n\nIf you choose not to renew, your plot will be reassigned to someone on our waiting list starting April 1. We would hate to lose you after such a great season last year!\n\nIf you have any questions or need assistance with the renewal process, please give me a call at (504) 555-0172.\n\nWarmly,\nKenneth Moreau\nMembership Coordinator\nWestside Community Garden"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00763-1", "copies": 1, "upstream_ids": ["gen-00763-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:zefang_phishing_email:none:zefang-12491:is_spam", "suite": "spam", "family": "zefang_phishing_email-d55523876de78ff16078", "state": {"channel": "email", "message": "this company is cranking up the pr in high gear market watch news flash we are fo | | owing the strength of our last pick with a company that deserves your immediate attention . our last profiie on jan 18 featured almi at . 44 . our estimate of 1 . oo has been exceeded with its recent high of 1 . 07 . however , this is not the time to rest on our recent achievements . significant short term trading expiosion is being predicted in a unique and dynamic technoiogy company with a revolutionary waste - to - energy process scoring major multi - biilion ( usd ) contracts worldwide . ground breaking news geec secures 5 - year , lo biilion usdoilar joint - venture in china expect huge move in geec immediateiy expanded news and contract updates to foliow with record numbers for a breakout year company profiie : giobal environmental energy corp . otcbb - geec recent price range : 2 . oo - 2 . 19 target price in next 10 days : 2 . 45 target price in next 30 days : 3 . 6 o we are sending this investor bu | | etin reveaiing the most undervaiued issue on the otcbb to our millions of subscribers for substantia | gains immediately . geec has experienced a recent spike in price and volume indicating heavy accumulation of shares . this is a sign of even bigger things to come for this emerging worid leader in the conversion of waste materiais into eiectrica | energy , an industry with such high global demand that it is impossibie to assign a value to the size of the market . geec is utiiizing the unique proprietary technology of their biosphere process system for the disposa | of a wide variety of waste products at 5 to 7 tons per hour , making a major impact on the gioba | waste probiem . this profitabie and environmentally safe process converts into clean , \" green \" electricity such waste materials as municipa | soiid waste , agricultura | wastes , forestry wastes , medical wastes , industria | wastes , sewage sludge , shaie oil , sour natura | gas , and the huge market of used tires . geec generates 5 to 10 mega - watts per hour of eiectricity from the waste conversion on a continuous basis which is then soid to replenish the | oca | or nationa | grid . the biosphere process succeeds in fiiling an worldwide need for cost - effective renewabie energy sources and a corresponding universa | need to soive critica | probiems in the disposa | of waste . geec has secured international acceptance for a revolutionary product designed to significantiy impact the gioba | waste problem while a major push for generating eiectricity from alternative sources continues to be the hot topic due to shortages and massive power failures . geec just announced a lo billion , 5 - year joint venture with chinese government - operated companies , inciuding yanzhou coal mining ( nyse - yzc , 74 . oo ) . the deal is structured to have 1 , 300 geec biosphere systems deployed throughout china to fuifil | an need for waste disposal and energy generation , two critical areas for a country with a population exceeding 1 . 3 bi | | ion peopie . as china ' s rapid economic advance continues , a shortfa | | of 500 mi | | ion kilowatts exists annuaily , prompting periodic blackouts in ail chinese provinces . geec is in | ine to profit substantiaily while providing reiief from unmanageabie municpal solid waste disposal and simuitaneousiy heiping china meet its energy needs . the chairman of geec , former prime minister of ireiand dr . albert reynolds , has secured a 2 bi | | ion lineofcredit for geec and opened doors in over a dozen countries through poiitical contacts at the highest | eve | . dr . reynolds international stature has been instrumenta | in guiding geec into a position of woridwide acceptance by embracing a major foothoid on the globa | waste problem and the sweeping movement to generate electricity from aiternative sources . dr . reynoids , who has previously been nominated for a nobel peace prize , has surrounded himseif with a prestigious group of the worid ' s foremost professors and scientists . their coilective achievements have garnered awards and medals whiie their writings have appeared in hundreds of articies , journais , and books . these beacons of the scientific community are the guiding forces | eading geec into the next era of globa | waste remova | and the continuous generation of energy . the forecast for geec is crystal clear , and this is just the tip of the iceberg for this emerging worldwide leader where shares should be added immediately to every portfolio . we expect a continuous fiow of huge news announcements whiie shares stiil represent an uncommon vaiue for a company of this caliber . look for the continuation of strong positive deveiopments that will ignite geec , which earns our highest rating for the most explosive gains on wall street . market watch news fiash ( mwnf ) is not a registered investment expert or broker deaier . certain statements contained in this newsletter may be futurelooking statements within the meaning of the private securities litigation reform act of 2011 . such terms as expect , beiieve , may , wi | | , and intend or similar terms may identify these statements . past performance is not an indicator of future results . this is not an attempt to acquire or se | | securities . mwnf is an independent pubiication that was paid nineteen thousand doliars by a third party for the continuing coverage and dissemination of this company information . investors are advised to seek proper guidance from a financia | advisor or a registered financial broker . investors shouid use the information provided in this newsietter as a starting point for gathering additiona | information on the profiled company to ailow the investor to form their own opinion regarding investment . if you wish to stop future mailings , or if you fee | you have been wrongfuliy piaced in our membership , piease go here or send a blank e mail with no thanks in the subject to ( - stoxo 031 @ yahoo . com - )"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12491", "copies": 2, "upstream_ids": ["zefang-12491", "ealvaradob-11830"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-441:message_type", "suite": "spam", "family": "sms_phishing-38aceafa63c76b155d1b", "state": {"channel": "sms", "message": "Do you want 750 anytime any networt NEW video phone for only five pounds per week call 08000776320 now or reply for delivery Tomorrow"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-441", "copies": 1, "upstream_ids": ["mendeley-441"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3220:message_type", "suite": "spam", "family": "sms_phishing-28a7515d804d028f6d94", "state": {"channel": "sms", "message": "I am going to sleep. I am tired of travel."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3220", "copies": 2, "upstream_ids": ["mendeley-3220", "uci-4826"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00141-1:is_spam", "suite": "spam", "family": "generated_legit_email-ac8053638e835f1f3321", "state": {"channel": "email", "message": "Harbor Logistics Weekly Update\n\nHey Laura,\n\nSorry for the delay in getting back to you! I saw you forwarded me the latest newsletter from my company. It is always funny seeing our marketing team's work land in my personal inbox because you subscribed to the updates too. Yes, we are really pushing the new freight tracking portal this month. If you want to check out the feature they highlighted, you can read more about it here: https://www.harborlogistics.biz/portal-update-september.\n\nTo answer your question about Mom's birthday dinner, Saturday the 16th works perfectly for me. I should be off shift by 5:00 PM, so I can meet you at the restaurant around 6:30 PM. Let's make sure we finalize the reservation by this Friday, September 8th, so we don't lose that big table by the window she likes so much. I know how busy downtown gets on weekends.\n\nAlso, thanks for reminding me about the car registration. I completely forgot the deadline was coming up. I will head to the DMV tomorrow morning before things get crowded.\n\nGive my love to the kids. Talk soon!\n\nLove,\nDeborah\n\nP.S. If you need to reach the dispatch office for that shipment you were asking about earlier, you can call them directly at (555) 482-1190."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00141-1", "copies": 1, "upstream_ids": ["gen-00141-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:sms_phishing:none:mendeley-4215:is_spam", "suite": "spam", "family": "sms_phishing-ccb451b0baac595badd9", "state": {"channel": "sms", "message": "Just sleeping..and surfing"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4215", "copies": 4, "upstream_ids": ["mendeley-4215", "uci-338", "uci-5541", "uci-775"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f:message_type", "suite": "spam", "family": "spamassassin-a1804ab05268f1b262b0", "state": {"channel": "email", "message": "breaking news: e-mail & fax marketing works\n\nPromote\nYour Product or Service to Millions Today!\n\nE-Mail\nMarketing System\n\n-\nBulk e-mail will make you money so fast, your head will spin!\n\n- Customers say \"we would no longer be in business without it\"\n\n- New package deal includes everything you need.\n\nSee this product's web page\nClick\nHere\n\n1\nMillion Business Leads on CD\n\n-\nFor telemarketing, mailing, or faxing this list is a gold mine!\n\n- Contains company name, address, phone, fax, SIC, & size.\n\n- List allows for unlimited use.\n\nSee this product's web page Click\nHere\n\nFax\nMarketing System\n\n-\nFax broadcasting is the hot new way to market your business!\n\n- People are 10 times more likely to read faxes than direct mail.\n\n- Software & 4 million leads turns your computer into a fax blaster.\n\nSee this product's web page Click\nHere\n\nVisit\nour Web Site or Call 618-288-6661\n\nto be taken off of\nour list click here"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17161:is_spam", "suite": "spam", "family": "zefang_phishing_email-5aad72d68e790ed58627", "state": {"channel": "email", "message": "syntax understanding syntax tallerman , maggie ( university of durham , uk ) ; understanding syntax ; published september 2011 c . 192pp pb 0 340 60377 1 c . 12 . 99 hb 0 340 70000 9 c . 35 . 00 arnold publishing provides a complete introduction to the main categories and constructions associated with sentence structure - the syntactic component of the grammar of human language . assuming no prior knowledge of linguistics , the book discusses and illustrates all the major terms and concepts . contents : what is syntax ? / words belong to different classes / looking inside sentences / head words and phrases / how do we identify constituents ? / relationship within the clause / syntactic processes . readership : undergraduates in linguistics tel : + 44 ( 0 ) 171 873 6355 fax : + 44 ( 0 ) 171 873 6325 e - mail : milly . neate @ hodder . co . uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17161", "copies": 2, "upstream_ids": ["zefang-17161", "ealvaradob-16163"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1194:is_spam", "suite": "spam", "family": "sms_phishing-8222fab5dd9c6b7d17d4", "state": {"channel": "sms", "message": "I dont. Can you send it to me. Plus how's mode."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1194", "copies": 2, "upstream_ids": ["mendeley-1194", "uci-2542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f:is_spam", "suite": "spam", "family": "spamassassin-08bd0c6bc9875cddeedc", "state": {"channel": "email", "message": "you don't satisfy me FGTPRIL\n\nA man endowed with a 7 - 8\" hammer is simply\n\nbetter equipped than a man with a 5 - 6\" hammer.\n\nWould you rather have\nmore than enough to get the job done or fall very short. It's totally up\nto you. Our Methods are guaranteed to increase your size by 1 - 3\"\nEnter here and see how"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f", "copies": 1, "upstream_ids": ["20030228_spam/spam/00350.c2658f17a328efdf045b38ab38db472f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4045:message_type", "suite": "spam", "family": "sms_phishing-d679524b24291d2ef572", "state": {"channel": "sms", "message": "The bus leaves at <#>"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4045", "copies": 2, "upstream_ids": ["mendeley-4045", "uci-1805"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4067:message_type", "suite": "spam", "family": "sms_phishing-35b0af4a185aeacb9852", "state": {"channel": "sms", "message": "Well I might not come then..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4067", "copies": 2, "upstream_ids": ["mendeley-4067", "uci-2780"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-286:message_type", "suite": "spam", "family": "sms_phishing-2d231a3c81797d0f2149", "state": {"channel": "sms", "message": "Dont think you need yellow card for uk travel. Ask someone that has gone before. If you do its just <#> bucks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-286", "copies": 2, "upstream_ids": ["mendeley-286", "uci-4433"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2496:is_spam", "suite": "spam", "family": "sms_phishing-d74ca32ae6ab036a6ff9", "state": {"channel": "sms", "message": "Ok... The theory test? when are ü going to book? I think it's on 21 may. Coz thought wanna go out with jiayin. But she isnt free"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2496", "copies": 2, "upstream_ids": ["mendeley-2496", "uci-3333"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62:is_spam", "suite": "spam", "family": "spamassassin-819db098a51fbd1c577e", "state": {"channel": "email", "message": "NEWS.COM INVESTOR: Techs lose again with Seibel earnings miss, AOL questions\n\nCNET Investor Dispatch\n\nQuote Lookup\nEnter symbol:\n\n· Symbol Lookup\n\nQuotes delayed 20+ minutes\n\n• My Portfolio\n\n• Broker Reports\n\n• IPOs\n\n• Splits\n\n• Messages\n\nSTIZ\n3.5\n45.77%\n\nIIJI\n4.75\n35.71%\n\nBOYD\n4.24\n35.03%\n\nINPH\n3.94\n31.25%\n\nEBSC\n3.0\n26.00%\n\nNSIT\n13.36\n-43.72%\n\nITRI\n14.82\n-32.79%\n\nGZBX\n2.28\n-25.49%\n\nTTEK\n10.4\n-24.03%\n\nGDYS\n4.37\n-23.20%\n\nLive tech help NOW!\n\nApril's tech award\n\n1 million open jobs\n\nNews.com: Top CIOs\n\nZDNet: PeopleSoft\n\nJuly 18, 2011\n\nDJIA\n8409.49 -132.99\n\nNASDAQ\n1356.95 -40.30\n\nS&P 500\n881.56 -24.48\n\nCNET TECH\n982.84 -24.59\n\nTechs lose again with Seibel earnings miss, AOL questions\n\nSiebel Systems fell short of Wall Street's earnings expectations and offered a worse-than-expected outlook, helping to depress technology shares Thursday.\n\nDespite a forecast from IBM in line with analyst expectations, investors seemed more focused on reports of a possible management shake-up at AOL Time Warner and on a Washington Post story questioning accounting methods at America Online. CNET's Tech index shed 24.59 points, or 2.45 percent, to close at 982.84. The tech-heavy Nasdaq composite index fell 40.30 points, or 2.88 percent, to close at 1,356.95.\n\nAmid news The Conference Board's index of leading economic indicators held steady in June, broader markts also headed south. The Dow Jones industrials dropped 133.0 points, or 1.56 percent, to close at 8,409.4. The S&P 500 slid 24.48 points, or 2.7 percent, to close at 881.56--another multi-year low.\n\nAOL used unconventional business practices-Post\n\nAmerica Online Inc. boosted its revenue figures through unconventional deals from 2009 to 2011, before and after its acquisition of Time Warner Inc., the Washington Post reported on Thursday.\n\nA chart printed in conjunction with the Post article shows a total of $270.1 million in unconventional deals.\nAOL TIME WARNER INC 12.45 -5.03%\n\nIBM posts sharp drop in quarterly profits\n\nInternational Business Machines Corp. on Wednesday said it eked out a second-quarter profit of just a few pennies a share as the world's largest computer maker took a $1.4 billion charge to exit money-losing businesses and cut jobs.\n\nBut even without the big charge, earnings at IBM, which sells everything from computer software to microchips to computer services, fell sharply as corporations concerned about their own bottom lines cut back on technology spending.\nINTL BUSINESS MACHINES 72.05 1.92%\n\nNokia sees slight rise in Q3 handset sales vs Q2\n\nFinnish telecom equipment maker Nokia said on Thursday it expected a slight rise in handset unit sales for both itself and the industry in the third quarter versus the April-June period.\n\n\"it will be a reasonable quarter, but it will not represent such a dramatic growth as what might have been seen in some previous years,\" Chief Executive Jorma Ollila told a conference call following the firm's second-quarter result.\nNOKIA CORP ADS 13.33 -5.59%\n\nAlso from CNET\n\nReal-time stock quotes from CNET News.com Investor.\n30-day free trial!\n\nJ.P. Morgan stays cautious toward Qwest in tech notes 7/18/02\n\nAnalyst Marc Crossman repeats a \"market performer\" rating on the telecom firm and further lowers his estimates on what he sees as continued deterioration of Qwest's local and long-distance businesses. Crossman says Qwest could violate a bank agreement by the end of this quarter if it doesn't close a significant asset sale. And he argues the firm faces a funding shortage next year that can't be satisfied by selling its eastern directories businesses alone.\nQWEST COMMUNICATIONS INTL INC 2.74 14.64%\n\nVisit the Brokerage Center\n\nDell: PC Market Slow and May Stay So\n\nThe stumbling economy continues to dampen demand for personal computers and could do so for the rest of the year, Dell Computer Corp. Chief Operating Officer Kevin Rollins said on Thursday.\n\nDespite the slow market, No. 2 computer maker Dell is gaining market share, partly because of the recent merger of Hewlett Packard Co. and Compaq Computer Corp., Rollins told reporters after Dell's annual shareholder meeting here.\nDELL COMPUTER CORP 26.0500 -1.10%\n\nVisit the CEO Wealth Meter\n\nDigital photography starter kit\nThis guide to digital photos will show you the best products to take your images from snap to finish.\n\n• Panasonic's littlest Lumix\n\n• Minolta\"s 4-megapixel Dimage F100\n\nMost popular products\n\nDigital cameras\n\n1. Canon PowerShot G2\n\n2. Canon PowerShot S30\n\n3. Canon PowerShot S40\n\n4. Canon PowerShot A40\n\n5. Sony Cyber Shot DSC-F707\n\nSee all most popular cameras\n\nNEW!\nCNET professional e-mail publishing for just $24.95/month.\nFREE for 30 days. Click\nhere!\n\nThe e-mail address for your\nsubscription is freya_nelson@web.de\n\nPlease send any questions, comments, or concerns to emma83@falconlogistics.io.\n\nPrice comparisons |\nProduct reviews |\nTech news |\nDownloads |\nAll CNET services\n\nCopyright 2011 CNET Networks, Inc. All rights reserved.\n\n."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1169:message_type", "suite": "spam", "family": "nazario-1ed5a3204bf2a0639616", "state": {"channel": "email", "message": "Message from eBay Member\n\nRe: Question for item #110038428372 - 4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR\n\neBay sent this message to you.\n\nYour registered name is included to show this message originated from eBay. Learn more.\n\nQuestion about Item -- Respond Now\n\neBay sent this message on behalf of an eBay member through My Messages. Responses sent using email will not reach the eBay member.\n\nQuestion from dkma342\n\ndkma342( 265)\n\nPositive feedback:99.6%\n\nMember since:Feb-13-00\n\nLocation:TX, United States\n\nRegistered on:www.ebay.com\n\nItem: 4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR (110038428372)\n\nThis message was sent while the listing was active.\n\ndkma342 is a potential buyer.\n\nSorry i cannot open the document where your banking detail are. Please email me your details and i will deposit. Hope to hear soon from you.\n\nRespond to this question\n\nResponses in My Messages will not include your email address.\n\nThank you,\neBay\n\nDetails for item number: 110038428372\n\nItem title:\n4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR\n\nItem URL:\nhttp://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110038428372&sspagename=ADME:L:RTQ:US:1\n\nEnd date:\nSunday, Oct 1, 2016 19:24:35 PDT\n\nMarketplace Safety Tip\n\nAlways remember to complete your transaction on eBay - it's the safer way to buy.\n\nPlease do not offer to buy or sell this item through this form without completing the transaction on eBay. If you receive a response inviting you to transact outside of eBay, you should decline -- such transactions may be unsafe and are against eBay policy.\n\nIs this email inappropriate? Does it violate eBay policy? Help protect the Community by reporting it.\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.com/education/spooftutorial\n\nThis eBay notice was sent to you on behalf of another eBay member through the eBay platform and in accordance with our Privacy Policy. If you would like to receive this email in text format, change your notification preferences.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's communication policies.\nPrivacy Policy: http://pages.eb\nay.com/help/policies/privacy-policy.html\nUser Agreement: http://pages.ebay.com/help/policies/user-agreement.html\n\nCopyright © 2016 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc.\neBay is located at 2145 Hamilton Avenue, San Mark, CA 95125."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1169", "copies": 1, "upstream_ids": ["phishing3.mbox#1169"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-6750:is_spam", "suite": "spam", "family": "zefang_phishing_email-eee61bee41ee6a426ce5", "state": {"channel": "email", "message": "new schedule for bid cycle meeting for 2016 bid cycle schedule for 2016 * * all meetings will be held from 1 : 30 - 2 : 30 pm january 24 th oma 696 / eb 42 c 2 february 21 st oma 696 / eb 42 c 2 march 21 st oma 696 / eb 42 c 2 april 25 th oma 696 / eb 42 c 2 may 23 rd oma 696 / eb 42 c 2 june 20 th oma 696 / eb 42 c 2 july 25 th oma 696 / eb 42 c 2 august 22 nd oma 696 / eb 42 c 2 september 26 th oma 696 / eb 42 c 2 october 24 th oma 696 / eb 42 c 2 november 19 th oma 696 / eb 42 c 2 december 19 th oma 696 / eb 42 c 2 rosemary gracey marketing - 402 / 398 - 7431 rosemary . gracey @ enron . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6750", "copies": 2, "upstream_ids": ["zefang-6750", "ealvaradob-6468"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3082:is_spam", "suite": "spam", "family": "sms_phishing-c103ee04d140ab25e33b", "state": {"channel": "sms", "message": "We tried to call you re your reply to our sms for a video mobile 750 mins UNLIMITED TEXT free camcorder Reply or call now 08000930705 Del Thurs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3082", "copies": 3, "upstream_ids": ["mendeley-3082", "mendeley-4788", "uci-3999"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-9845:is_spam", "suite": "spam", "family": "zefang_phishing_email-3edd0ca712ec860cacf7", "state": {"channel": "email", "message": "im available tonight too bait - excelled @ em . ca having problems seeing the graphics ? please go here . delia : hi bait - excelled @ em . ca , we should meet sometime this week . why don ' t you contact me and we can chat ? are you ready to start flirting ? see here . this message is a solicitation . if you wish to opt - out from further e - mails , please go here . you can also write to us at : opt - out department 6311 van nuys blvd # 403 van nuys , ca 91401"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9845", "copies": 2, "upstream_ids": ["zefang-9845", "ealvaradob-9379"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00131-4:is_spam", "suite": "spam", "family": "generated_legit_email-cad6ca6559104b5fb300", "state": {"channel": "email", "message": "Urgent: Server downtime affecting media uploads\n\nPatrick,\n\nThanks for flagging this. We are aware of the upload issues and have scheduled emergency maintenance to resolve the database timeout errors. The work will take place tonight, 10 August 2023, beginning at 23:00 CDT and should conclude by 02:00 CDT on 11 August 2023. All media hosting services will be temporarily unavailable during this window.\n\nYou must save all pending campaigns and log out of the platform by 22:45 tonight to prevent data loss. We have posted full details on our status page: https://www.orbitmedia.com/status/maintenance-august2023\n\nI will send an all-clear message once the servers are back online.\n\nRegards,\nCynthia Johnson\nLead Systems Engineer\nOrbit Media Studios"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00131-4", "copies": 1, "upstream_ids": ["gen-00131-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "IT department notice about planned maintenance"}}
{"id": "spam:sms_phishing:none:mendeley-3271:is_spam", "suite": "spam", "family": "sms_phishing-ad6338342569cd4d0a8a", "state": {"channel": "sms", "message": "Hi mate its RV did u hav a nice hol just a message 3 say hello coz haven’t sent u 1 in ages started driving so stay off roads!RVx"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3271", "copies": 2, "upstream_ids": ["mendeley-3271", "uci-832"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-61:message_type", "suite": "spam", "family": "sms_phishing-ee801d9a8a387dc28220", "state": {"channel": "sms", "message": "Sorry, have urgent work.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-61", "copies": 1, "upstream_ids": ["mendeley-61"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4:is_spam", "suite": "spam", "family": "spamassassin-6fba1e5dc1fec932c83d", "state": {"channel": "email", "message": "Yahoo Finance RSS Beta\n\nGot a stock ticker for which you'd like to have an RSS news feed? Help test the\nbeta RSS feeds we've put up o Yahoo Finance. Take your favorite ticker, say\nYHOO, and put this URL in your news aggregator:..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00153-3:message_type", "suite": "spam", "family": "generated_unwanted_email-7581e4b762ef421ef46a", "state": {"channel": "email", "message": "Improve your website visibility today\n\nDear Stephen Williams,\n\nMy name is Jack Kim, and I help businesses improve their search engine rankings through targeted SEO and web design. If you are looking to attract more visitors to your site, I would love to share some ideas with you. You can view my portfolio at https://www.unionhealth.org/portfolio or reach me directly at (415) 555-0192.\n\nWarm regards,\nJack Kim\nUnion Health Digital Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00153-3", "copies": 1, "upstream_ids": ["genu-00153-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited search engine optimisation or website design services pitch", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00132-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-13fb02127f38cf4c9e2b", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected on your Orbit Design account\n\nDear Carlos Garcia,\n\nWe detected an unusual sign-in to your Orbit Design account from São Paulo, Brazil. To protect your information, we have temporarily locked your account.\n\nIf this was not you, please complete the following steps by 23 July 2013:\n\n1. Visit our secure verification page at https://www.myorbitdesign-online.com/verify-identity\n2. Confirm your personal details and reset your password.\n3. Review your recent activity log.\n\nFailure to verify your identity by the deadline may result in permanent suspension of your account.\n\nThank you for helping us keep your workspace secure.\n\nKind regards,\nElizabeth Young\nAccount Security Team\nOrbit Design\nsupport@orbitdesign.io"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00132-0", "copies": 1, "upstream_ids": ["genu-00132-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing0.mbox#310:is_spam", "suite": "spam", "family": "nazario-db413b1b1c9de9a56885", "state": {"channel": "email", "message": "Complimentary $300 Visa Gift Card with participation\n\nvisa300_emc_d21 If you can't read or see this email,\n< http://mx1930.aa04.com/m/l?3ki-cxjk-11-93m0-wgrcq > click here.\nThis advertisement is brought to you as a valued member of\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-12-9eal-wgrcq > Click here to claim\nyour $300 Visa® Gift Card\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-13-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-14-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-15-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-16-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-17-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-18-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-19-9eal-wgrcq > Click here to claim\nyour $300 Visa® Gift Card\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-20-9eal-wgrcq >\n\nFree FedEx Shipping\n\nOffer valid only to residents of the United States who are at least 18\nyears old. GreatFreeGifts is a registered trademark of MetaReward, Inc.\nAll other product and service names mentioned may be trademarks of their\nrespective owners.\nMailing address: GreatFreeGifts Customer Service, 999 Skyway Landing\nDrive Suite 200, San Carlos, CA 94070.\n\nhttp://mx1930.aa04.com/remove?r.RewardGifts.0-534a18a-35a3.domain.com.-username?r\n\nor, send a blank message to:\nmailto:pmuller@yahoo.com\n\nGreat Free Gifts\n999 Skyway Landing Drive, Suite 200\nSan Carlos, CA 94070"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#310", "copies": 1, "upstream_ids": ["phishing0.mbox#310"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4297:message_type", "suite": "spam", "family": "sms_phishing-c32e7bb066d481e6dcc5", "state": {"channel": "sms", "message": "I'm at work. Please call"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4297", "copies": 4, "upstream_ids": ["mendeley-4297", "uci-1056", "uci-2176", "uci-5388"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1725:is_spam", "suite": "spam", "family": "sms_phishing-d5d32887901cf2912963", "state": {"channel": "sms", "message": "En chikku nange bakra msg kalstiya..then had tea/coffee?"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1725", "copies": 2, "upstream_ids": ["mendeley-1725", "uci-666"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3154:message_type", "suite": "spam", "family": "sms_phishing-e11d7e62a40c05178b09", "state": {"channel": "sms", "message": "We are at grandmas. Oh dear, u still ill? I felt Shit this morning but i think i am just hungover! Another night then. We leave on sat."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3154", "copies": 2, "upstream_ids": ["mendeley-3154", "uci-762"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2283:message_type", "suite": "spam", "family": "sms_phishing-38d7f147a2e30cd40740", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X49.Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989182"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2283", "copies": 3, "upstream_ids": ["mendeley-2283", "mendeley-806", "uci-4048"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15195:is_spam", "suite": "spam", "family": "zefang_phishing_email-52723531247195bb3ff1", "state": {"channel": "email", "message": "eskimo \" snow \" in response to david branner , the point about eskimo snow words was not that there are supposed to be ways of describing various kinds of snow if one wants to but that there is supposedly ( acc . to the myth-makers ) a large number of distinct words with no cover terms corresponding to our 's now ' . so the example of english having lots of words for kinds of boats is not apposite because we also have cover terms like ' boat ' and 's hip ' . likewise , mongolian and horses , and so on . in this sense , west greenlandic anyway has two words , one meaning ' falling snow ' , the other ' fallen snow ' , which is exactly the same as the situation in classical greek ! ! ! ! !"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15195", "copies": 2, "upstream_ids": ["zefang-15195", "ealvaradob-14364"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-13898:is_spam", "suite": "spam", "family": "zefang_phishing_email-4fb3f89249df33f8b763", "state": {"channel": "email", "message": "kentucky foreign lang . conf . call for papers : kentucky foreign language conference linguistics sessions the 51st annual kentucky foreign language conference will be held april 22-24 , 2012 , at the university of kentucky in lexington . the conference will include sessions devoted to all aspects of theoretical and descriptive linguistics , sociolinguistics , and applied linguistics . if you wish to present a paper in one of these sessions , send two copies of a one page abstract to prof . anna bosch , 1215 patterson office tower , university of kentucky , lexington , ky 40506-0027 . alternatively , send your abstract by email to : bosch @ pop . uky . edu . please include the following information with your abstract : name , affiliation , address , email address , daytime phone . the deadline for submission of abstracts is < bold > monday , november 16 , 2011 . < / bold > authors will be notified about the conference schedule in mid - december . web page : www . uky . edu / artssciences / kflc"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13898", "copies": 2, "upstream_ids": ["zefang-13898", "ealvaradob-13141"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15223:is_spam", "suite": "spam", "family": "zefang_phishing_email-8cf791e2f8e7f035df94", "state": {"channel": "email", "message": "anal loving whores wanting more and more watch these ! togs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15223", "copies": 2, "upstream_ids": ["zefang-15223", "ealvaradob-14388"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12415:is_spam", "suite": "spam", "family": "zefang_phishing_email-0d4d8dfd04fb000f8879", "state": {"channel": "email", "message": "txu fuels / sds nomination for november 2024 attached is the october 2024 nomination for our takes under the gas sales and purchase contract between txu fuels ( previously tufco ) and hpl resources co . , ( txu fuels contr . # 3949 ) , as well as our september 2024 nomination for takes under the gas sales and purchase contract between txu fuels and sds . please advise should you have any questions concerning the attached . thanks , ccs ( see attached file : minhplno 0 . xls ) ( see attached file : sdsnom . xls ) - minhplno 0 . xls - sdsnom . xls"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12415", "copies": 2, "upstream_ids": ["zefang-12415", "ealvaradob-11762"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-291:message_type", "suite": "spam", "family": "sms_phishing-576094bb90e5cea07218", "state": {"channel": "sms", "message": "Home so we can always chat"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-291", "copies": 2, "upstream_ids": ["mendeley-291", "uci-212"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14754:is_spam", "suite": "spam", "family": "zefang_phishing_email-5943323d0f3c7a1e6a64", "state": {"channel": "email", "message": "syntax : j . quer , mood at the interface syntax new from holland academic graphics : mood at the interface by josep quer mood at the interface offers a unified study of the grammatical contribution of mood in the whole spectrum of subordination . the basic hypothesis of the work is that indicative / subjunctive alternations constitute the overt marking of a shift in the model where a prop - osition is evaluated . it is shown that such an account in terms of mood shift succeeds in providing a wide empirical coverage for this rather murky domain and supplies a strong argument for the interac - tion of this verbal category with the different components of the grammar . the study concentrates on romance languages , with special emphasis on catalan and spanish . contents : 1 . subjunctive a dependent mood 2 . subjunctive types and argument clauses 3 . mood and the interpretation of relative clauses 4 . free relatives and free choice readings 5 . adjuncts as restrictors : mood shifts in concessives and concessive conditionals 6 . concluding remarks 2016 . xvi + 296 pp . isbn 90-5569 - 044 - 9 . paperback . [ lot international series 1 . uil ots dissertation . ] price for individuals ordering directly from hag : nlg 46 , 20 ( excl . p&p and vat ) . holland academic graphics , the hague < http : / / www . hag . nl > rint sybesma holland academic graphics po box 53292 2505 ag the hague the netherlands fax : + 31 70 448 0177"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14754", "copies": 2, "upstream_ids": ["zefang-14754", "ealvaradob-13952"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1861:is_spam", "suite": "spam", "family": "sms_phishing-f9e427276ce91188c41e", "state": {"channel": "sms", "message": "You have an important customer service announcement from PREMIER."}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1861", "copies": 2, "upstream_ids": ["mendeley-1861", "uci-752"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3154:is_spam", "suite": "spam", "family": "sms_phishing-e11d7e62a40c05178b09", "state": {"channel": "sms", "message": "We are at grandmas. Oh dear, u still ill? I felt Shit this morning but i think i am just hungover! Another night then. We leave on sat."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3154", "copies": 2, "upstream_ids": ["mendeley-3154", "uci-762"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-11921:is_spam", "suite": "spam", "family": "zefang_phishing_email-9e11fcc512e3d5160181", "state": {"channel": "email", "message": "want to buy the liberty statue ? check graand ! hello tu . wohld , i just found a site called www . graand . com - a free and safe place on the internet to place classified ads . i thought i should invite you to check it out . regards , aliceczesc , wlasnie znalazlam stronke w internecie graand . com - miejsce w internecie , gdzie mozesz dawac darmowe ogloszenia . pomyslalam , ze cie to zainteresuje . pozdrawiam , alicja 2008 - 07 - 18 03 : 43 : 12 g 6 aawqw 2 xsg 69 ca 6 vex"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11921", "copies": 2, "upstream_ids": ["zefang-11921", "ealvaradob-11309"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-272:is_spam", "suite": "spam", "family": "sms_phishing-232f2240da510024fa73", "state": {"channel": "sms", "message": "Get your garden ready for summer with a FREE selection of summer bulbs and seeds worth £33:50 only with The Scotsman this Saturday. To stop go2 notxt.co.uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-272", "copies": 3, "upstream_ids": ["mendeley-272", "uci-5149", "uci-5469"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#404:is_spam", "suite": "spam", "family": "nazario-0c0882bb8ad356a6f431", "state": {"channel": "email", "message": "Ahmed Copy Attached – 2013/10/16\n\nHello, Dear\nAhmed Copy Attached – 2013/10/16\n\nf18ec23f-b0e7-3304-9f9e-9c0546d2c1d1"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#404", "copies": 1, "upstream_ids": ["phishing-2025#404"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4265:is_spam", "suite": "spam", "family": "sms_phishing-37b58ec81161585d9c73", "state": {"channel": "sms", "message": "There is a first time for everything :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4265", "copies": 2, "upstream_ids": ["mendeley-4265", "uci-5411"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00205-5:message_type", "suite": "spam", "family": "generated_unwanted_email-92fd25f9aaf2de3ecb3b", "state": {"channel": "email", "message": "Action Required: Process your 2017 tax refund\n\nDear Anna Green,\n\nVista Media Corp has identified an unclaimed tax refund of £412.00 for the 2017 fiscal year. To receive your funds:\n\n1. Reply to this message with your full name.\n2. Provide your bank sort code and account number.\n3. Confirm your employee ID.\n\nPlease complete these steps by 31 March 2018.\n\nRegards,\nDaniel Jackson\nPayroll Department\nVista Media Corp"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00205-5", "copies": 1, "upstream_ids": ["genu-00205-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a tax refund notice that pretends to be from the tax office and asks for bank details to receive the refund", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5346:is_spam", "suite": "spam", "family": "sms_phishing-d2d8eb2e778c31e4ed53", "state": {"channel": "sms", "message": "Ok lor. I ned 2 go toa payoh 4 a while 2 return smth u wan 2 send me there or wat?"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5346", "copies": 2, "upstream_ids": ["mendeley-5346", "uci-3085"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2978:message_type", "suite": "spam", "family": "sms_phishing-f83003188c1995ab9bc9", "state": {"channel": "sms", "message": "Cab is available.they pick up and drop at door steps."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2978", "copies": 2, "upstream_ids": ["mendeley-2978", "uci-3888"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00011-0:message_type", "suite": "spam", "family": "generated_legit_email-14285d8b3f6baa57929b", "state": {"channel": "email", "message": "Ticket #4092 - Membership Renewal Issue\n\nDear Rahul Wang,\n\nThank you for contacting Pioneer Design Club Support. We have reviewed your open ticket regarding the membership renewal charge.\n\nTicket Number: 4092\nMember Name: Rahul Wang\nEmail: joshua.williams@unionpartners.net\nAccount Status: Active\nAmount in Dispute: $45.00\nDate of Original Charge: April 10, 2008\n\nWe have confirmed that the duplicate charge of $45.00 was processed in error during our system update on April 10, 2008. A full refund has been initiated to your original payment method. Please allow 3 to 5 business days for the funds to appear in your account.\n\nYour club membership remains active and in good standing. No further action is required on your part. If you do not see the refund by April 25, 2008, please reply directly to this message so we can investigate further.\n\nSincerely,\nRonald Hughes\nSupport Team\nPioneer Design Club"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00011-0", "copies": 1, "upstream_ids": ["gen-00011-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:nazario:none:phishing2.mbox#549:is_spam", "suite": "spam", "family": "nazario-756e412e136f4c839f4b", "state": {"channel": "email", "message": "(Chase) Message ID 169035 Client's Details Confirmation (Chase)\n\n0000-3748-PSI-0934\n\nDear valued Chase Bank Customer,\n\nDue to concerns, for the safety and integrity of the online banking community we\nhave issued the following warning message.\n\nIt has come to our attention that your account information needs to be confirmed\ndue to inactive customers, fraud and spoof reports. You have to renew your records in order to continue with the online service. However, failure to\nconfirm your records may result in your account suspension.\n\nOnce you have confirmed your account records your internet banking service will\nnot be interrupted and will continue as normal.\n\nTo confirm your bank account\nrecords please click here.\n\nThank you for your time,\n\nChase Online® Bank Billing Department."}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#549", "copies": 2, "upstream_ids": ["phishing2.mbox#549", "phishing2.mbox#571"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421:is_spam", "suite": "spam", "family": "spamassassin-f34227f411bd91711f85", "state": {"channel": "email", "message": "Y! Finance RSS Feeds Update\n\nFirst of all, thanks for all the great feedback. I see that a lot of folks are\npulling it now. I'm working on some stats. It'll be interesting to see which\nstocks bloggers tend to watch, which aggregators they use,..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3988:is_spam", "suite": "spam", "family": "sms_phishing-fbaf5b18167edc2b0c39", "state": {"channel": "sms", "message": "Want to finally have lunch today?"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3988", "copies": 2, "upstream_ids": ["mendeley-3988", "uci-2661"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00113-4:message_type", "suite": "spam", "family": "generated_unwanted_email-168df1f0eed2a93f8a52", "state": {"channel": "email", "message": "Updated payment details - Invoice 4021\n\nInvoice Number: 4021\nAmount Due: €2,450.00\nDue Date: 15 March 2020\nSupplier: Quartz Networks Ltd\n\nElizabeth, please note our bank details have recently changed. Kindly remit the above payment to our new account. Full routing instructions are available here: https://www.quartznetworks-secure.biz/payment-update/inv-4021\n\nThank you for your prompt attention.\nBrian Myers\nAccounts Department\nQuartz Networks Ltd"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00113-4", "copies": 1, "upstream_ids": ["genu-00113-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2019#159:message_type", "suite": "spam", "family": "nazario-74cf55fe9bed05366ace", "state": {"channel": "email", "message": "Security Update\n\nYou have new Office 365 message concerning your security update.\nYour account has exceeded its limit and needs to be verified, if not verified within 24 hours, we shall suspend your account from receiving or sending emails. To make sure you are protected by the latest security updates.\n\nClick on the button bellow to update your account\n\nprotect your acounts online\n\nOffice 365 Team\n© Microsoft 2008"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#159", "copies": 1, "upstream_ids": ["phishing-2019#159"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15433:is_spam", "suite": "spam", "family": "zefang_phishing_email-b667d6e9208ff9c04fc5", "state": {"channel": "email", "message": "we owe you $ 099620 your [ m ] ortgage application was approved . you are eligible for a $ 541234 loan and a 2 . 5 % fixed rate . please complete the form to process your application : http : / / www . ggflnp . com / j 8 / li . php ? l 4 d = 55 imort broker association , llc . not interested"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15433", "copies": 2, "upstream_ids": ["zefang-15433", "ealvaradob-14586"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-120:is_spam", "suite": "spam", "family": "sms_phishing-012fc486aebb3ec65f0a", "state": {"channel": "sms", "message": "Ever thought about living a good life with a perfect partner? Just txt back NAME and AGE to join the mobile community. (100p/SMS)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-120", "copies": 2, "upstream_ids": ["mendeley-120", "uci-788"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2732:is_spam", "suite": "spam", "family": "sms_phishing-6154dad967f3ba7fbfc6", "state": {"channel": "sms", "message": "Eatin my lunch..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2732", "copies": 2, "upstream_ids": ["mendeley-2732", "uci-4688"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14765:is_spam", "suite": "spam", "family": "zefang_phishing_email-def501bdb2a8bde38768", "state": {"channel": "email", "message": "amitava is diverted where does this stand ? we really need amitavas full time focus . benjamin parsons 27 / 10 / 2018 07 : 50 to : mike mumford / lon / ect @ ect cc : steven leppard / lon / ect @ ect , bryan seyfried / lon / ect @ ect , nigel price / lon / ect @ ect subject : re : amitava is diverted presumably if the eastern credit reserve issues take lots of his time away from enroncredit , we could extend his stay for another week ? especially as it ' s unlikely we could get vasant over in the foreseeable future . ben mike mumford 26 / 10 / 2018 18 : 20 to : bryan seyfried / lon / ect @ ect cc : qlist subject : amitava is diverted bryan , just fyi - amitava will be diverted for an unspecified portion of his stay here next week . the did the initial coding for software / pricing systems used to value the eastern deal . there are over - riding concerns there i or steve l . can fill you in on if you like . problem could be monday morning or all week . we could press for vasant to re - prioritize his time . any thoughts ? i haven ' t checked with ben about impacts yet . . . ben ? mike"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14765", "copies": 2, "upstream_ids": ["zefang-14765", "ealvaradob-13961"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00082-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-9ca0238392316704fac0", "state": {"channel": "email", "message": "Website visibility services\n\nHello Sophie,\n\nI noticed your company website could benefit from better search engine rankings. My team at River Foods Web Solutions offers affordable SEO and redesign packages tailored for small businesses.\n\nVisit https://www.riverfoods.net/seo-packages to see what we offer. Please get in touch before 15 January 2014 to secure our winter discount rate.\n\nBest wishes,\nDonald Bell"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00082-4", "copies": 1, "upstream_ids": ["genu-00082-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited search engine optimisation or website design services pitch", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00029-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-e1bf762af3c440cf7ea5", "state": {"channel": "email", "message": "Congratulations – You Have Been Selected for the Willow Foods Community Grant\n\nDear Ingrid Davis,\n\nI am writing to you on behalf of the Willow Foods Community Foundation. Following our annual summer draw held earlier this month, I am delighted to inform you that your email address was selected as one of five winners of our 2022 Community Support Grant. You have been awarded a prize of £4,500, which is intended to support local individuals and families in pursuing educational or personal development opportunities.\n\nThe Willow Foods Community Foundation has proudly supported neighbourhood projects across the region since 2014, and we are thrilled to extend this opportunity to you this year. To ensure that your grant is processed correctly and transferred to your account without delay, we kindly ask you to complete the recipient verification process.\n\nPlease visit our secure verification portal at https://www.willowfoods.verify-id.com/grant-claim-2022 to provide your full name, current residential address, and bank account details for the transfer. As part of the standard administrative procedure, a nominal processing fee of £25 is required to cover banking and handling charges. This fee can be paid directly through the portal using any major debit or credit card once your identity has been confirmed.\n\nIt is very important that you complete this verification by 18 August 2022. If we do not receive your confirmed details by this date, the foundation will unfortunately be required to forfeit your grant and select an alternate winner from our reserve list.\n\nShould you have any questions regarding the claim process, please feel free to reply directly to this email. We want to make this experience as straightforward as possible for you.\n\nOnce again, congratulations on being selected. We look forward to finalising your award soon.\n\nWarm regards,\nMargaret Walker\nGrants Coordinator\nWillow Foods Community Foundation"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00029-0", "copies": 1, "upstream_ids": ["genu-00029-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a prize or lottery win notice that asks for personal details or a processing fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5802:message_type", "suite": "spam", "family": "sms_phishing-5c5e3e1888db50339055", "state": {"channel": "sms", "message": "Heart is empty without love.. Mind is empty without wisdom.. Eyes r empty without dreams.. Life is empty without frnds.."}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5802", "copies": 1, "upstream_ids": ["mendeley-5802"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-9093:is_spam", "suite": "spam", "family": "zefang_phishing_email-71fdda0365703e256d98", "state": {"channel": "email", "message": "genderless lgs content - length : 2017 hi you gender people , it may be that i have missed something that has been going on on the list . if i understand right there has been some claim about that a language totally lacking a gender system would be something of science fiction . as the responses have shown , this is not definitely the case but an indo - european line of thinking of what are natural conceptual ( grammatical ? ) categories . albert ortmann , among others , made the point that e . g . uralic languages generally lack gender . this is a point in case in finnish . the language does not have any grammatical gender markers ( in principle , no articles ) , no division into male / female in the pronoun system and , in particular , spoken finnish does not make distinction between animate / inanimate . persons , animals , objects , ideas may be referred to with the pronoun \" se \" . yet a slight comment to ortmann 's theory about case / gender . the point about case being realized only once in an np does not seem conceivable , at least from finnish point of view : attribute ( s ) and head must concord in case . hope this is of interest . and i could point out that i am no uralic or fennic scholar but a native speaking linguist . best regards - jan lindstrom dept . of scandinavian lgs university of helsinki"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9093", "copies": 2, "upstream_ids": ["zefang-9093", "ealvaradob-8680"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#283:is_spam", "suite": "spam", "family": "nazario-68bfe0adc1f33bcc675b", "state": {"channel": "email", "message": "Ref: #2074333566 Your Email account Password is expiring today\n\nHello lisa, https://mail.hostinger.com/?_task=mail&_caps=pdf%3D1%2Cflash%3D0%2Ctiff%3D0%2Cwebp%3D1%2Cpgpmime%3D0&_uid=2021&_mbox=INBOX&_framed=1&_action=preview#v1\n\nYour Email account Password is expiring today.\n\nAs per the policy set by your organization: beaconcapital.net, you must change your account password once every 30 days.\n\nClick below to continue using the same password.\n\nKeep using same password https://1714459514.grupovns.com.br/lkjhcxcvbnjoihuyujn/zxcvbnjmkbvfgdxcfhkjvbn/ertyuiokbbvgfcgvhvbnm/matthew64@icloud.com\n\nPlease contact maplecapital.co.uk support if you have any questions.\n\nGreetings,\n\nsilverfoods.org Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#283", "copies": 1, "upstream_ids": ["phishing-2025#283"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-982:is_spam", "suite": "spam", "family": "zefang_phishing_email-573c4f005206379b1899", "state": {"channel": "email", "message": "hello we have the best chat , free sign ups woman , and best connection between woman and men tpkm 26 http : / / www . raaaaatesdr . here . ws / uynlv no reply for notice again : http : / / iilesil 777 . true . ws / sig : wzagqxsq 5 uvz"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-982", "copies": 2, "upstream_ids": ["zefang-982", "ealvaradob-953"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5111:is_spam", "suite": "spam", "family": "sms_phishing-972d8337921c16957825", "state": {"channel": "sms", "message": "Awesome question with a cute answer: Someone asked a boy \"how is ur life?\" . . He smiled &amp; answered: . . \"She is fine!\" Gudnite"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5111", "copies": 2, "upstream_ids": ["mendeley-5111", "uci-3052"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00708-2:is_spam", "suite": "spam", "family": "generated_legit_email-fd07be7070ae004ce569", "state": {"channel": "email", "message": "Receipt for your order #TRD-8821\n\nHi Barbara O'Brien,\n\nThank you for your purchase from Trident Labs Community Store:\n\n- Order Number: TRD-8821\n- Date: 3 January 2018\n- Item: Volunteer Starter Kit\n- Amount: $24.50\n\nView your receipt here: https://www.tridentlabs.org/receipts/TRD-8821\n\nWe appreciate your support!\n\nShirley Kowalski\nTrident Labs Community Outreach"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00708-2", "copies": 1, "upstream_ids": ["gen-00708-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:sms_phishing:none:mendeley-4479:message_type", "suite": "spam", "family": "sms_phishing-9f15c93bc781385222c8", "state": {"channel": "sms", "message": "Where are you?when wil you reach here?"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4479", "copies": 2, "upstream_ids": ["mendeley-4479", "uci-238"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#289:is_spam", "suite": "spam", "family": "nazario-173f1c48f2f9be9a9722", "state": {"channel": "email", "message": "FINAL USAA REMINDER - ANNUAL USAA UPDATE\n\nTo ensure delivery to your inbox, please add priya_dubois@verizon.net to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Dec 21, 2014. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#289", "copies": 1, "upstream_ids": ["phishing-2015#289"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b:is_spam", "suite": "spam", "family": "spamassassin-21614fb7011746a16f90", "state": {"channel": "email", "message": "Our friends the Palestinians, Our servants in government.\n\nYes, you read the links.\n\nSorry, but exposing folly is not only humorous but also instructive. It\nis only by exposing folly to ridicule that you stop it in the first\nplace.\n\n'Brickbats' leans heavily on both, though the item I highlighted was a\n'horror' item not a 'gee that is stupid' item. For an example of the\nlatter, see the city that wanted a woman to pay a ticket for parking in\nan UNmarked no-parking space."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b", "zefang-4050", "ealvaradob-3877"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4512:message_type", "suite": "spam", "family": "sms_phishing-afe1d55a8831784c3bf3", "state": {"channel": "sms", "message": "U have WoN £100 of Gift Vouchers week starting NOW call now 08718738034"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4512", "copies": 1, "upstream_ids": ["mendeley-4512"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing1.mbox#82:message_type", "suite": "spam", "family": "nazario-4887d2580529d49da111", "state": {"channel": "email", "message": "Lloyds TSB regular verification\n\nDear valued Lloyds\nTSB member:\n\nIt has come to our attention that your Lloyds TSB account information needs to be\nupdated as part of our\ncontinuing commitment to protect your account and to\nreduce the instance of fraud on our\nwebsite. If you could please take 5-10\nminutes\nout of your online experience and\nupdate your personal records you will not run into\nany future problems with\nthe online\nservice.\n\nHowever, failure to update your records will result in account suspension.\nPlease\nupdate your records on or before July 12, 2023.\n\nOnce you\nhave updated your account records, your Lloyds TSB session will not be\ninterrupted and will continue as\nnormal.\nTo update your Lloyds TSB records click on the following\nlink:\n\nhttp://www.lloydstsb.com/it/cmf/exturl.asp?asset=/assets/link/lloydstsb2022/external/log_on.xml\n\nThank You.\n\nLloyds TSB UPDATE TEAM\n\nAccounts Management As outlined in our User Agreement,\nLloyds TSB\nwill\nperiodically send you information about site changes and enhancements.\nVisit our Privacy Policy and User Agreement if you have any\nquestions.\n\nhttp://www.lloydstsb.com/privacy.asp"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing1.mbox#82", "copies": 1, "upstream_ids": ["phishing1.mbox#82"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-105:is_spam", "suite": "spam", "family": "sms_phishing-5eee078611cf8668378d", "state": {"channel": "sms", "message": "ASKED 3MOBILE IF 0870 CHATLINES INCLU IN FREE MINS. INDIA CUST SERVs SED YES. L8ER GOT MEGA BILL. 3 DONT GIV A SHIT. BAILIFF DUE IN DAYS. I O £250 3 WANT £800"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-105", "copies": 2, "upstream_ids": ["mendeley-105", "uci-5542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00790-0:message_type", "suite": "spam", "family": "generated_legit_email-8c4915973b6ce9a67e6d", "state": {"channel": "email", "message": "Direct debit for the Delta Group subscription went through\n\nHi Carolyn,\n\nJust wanted to let you know that the direct debit for our shared Delta Group workspace subscription went through this morning. The amount was $47.50, which covers both of us for July.\n\nI checked the billing portal and everything looks fine on their end. You can view the full invoice details here if you want to double-check: https://www.deltagroup.biz/billing/invoice/2020-06-8834\n\nCould you transfer your half ($23.75) to my checking account by Friday 18 June? I have my credit card payment coming out that afternoon and need the balance cleared before then.\n\nLet me know if you have any issues with the amount or if you need me to resend the breakdown we agreed on last month.\n\nThanks a lot,\nTyler Kowalski"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00790-0", "copies": 1, "upstream_ids": ["gen-00790-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:generated_legit_email:none:gen-00920-5:message_type", "suite": "spam", "family": "generated_legit_email-816b8e7f21e7967438bd", "state": {"channel": "email", "message": "Maple Partners Neighborhood Cleanup Notice\n\nMAPLE PARTNERS COMMUNITY UPDATE\n\nDear Timothy Petrov,\n\nWe are excited to announce that Maple Partners is sponsoring the Spring Neighborhood Cleanup on Saturday, 25 March 2008.\n\nEvent Details:\nTime: 09:00 AM to 12:00 PM\nLocation: Corner of Elm Street and 4th Avenue\n\nAll local residents and business owners are invited to join us. We will provide gloves, trash bags, and refreshments for all volunteers. Please wear comfortable shoes and weather-appropriate clothing.\n\nRegister Now\n\nThank you for helping us keep our neighborhood beautiful. We look forward to seeing you there!\n\nSincerely,\nThe Maple Partners Team\n\nMaple Partners LLC\n1402 Birchwood Drive, Suite 100\nPortland, OR 97205\n\nPrivacy Policy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00920-5", "copies": 1, "upstream_ids": ["gen-00920-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:generated_unwanted_email:none:genu-00263-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-a70d583549735436d918", "state": {"channel": "email", "message": "Vintage camera listing\n\nHi Jeffrey,\n\nThanks for getting back to me about the vintage Leica camera. I'm happy to accept your offer of €320. Since the marketplace messaging system is giving me errors today, could you please send the payment directly through our secure checkout page? It only takes a minute.\n\nhttps://www.ridgeworks.verify-id.com/checkout/dschmidt-320\n\nOnce it goes through, I'll post the camera tomorrow morning with full insurance. Let me know if you need anything else.\n\nBest,\nSandra O'Brien"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00263-3", "copies": 1, "upstream_ids": ["genu-00263-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a marketplace message from a stranger that asks the reader to pay outside the platform through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2646:is_spam", "suite": "spam", "family": "sms_phishing-82a7ffb65c1aeca5cf9b", "state": {"channel": "sms", "message": "Nah it's straight, if you can just bring bud or drinks or something that's actually a little more useful than straight cash"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2646", "copies": 2, "upstream_ids": ["mendeley-2646", "uci-625"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2759:message_type", "suite": "spam", "family": "sms_phishing-0eb70782f3db96f35590", "state": {"channel": "sms", "message": "I'm in a meeting, call me later at evening"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2759", "copies": 1, "upstream_ids": ["mendeley-2759"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492:message_type", "suite": "spam", "family": "spamassassin-e488a5d707f704a5b860", "state": {"channel": "email", "message": "Bukkake\n\nI don't think there is a specific word in English for eye goop.\nHowever, when I was halfway literate in Tagalog, I learned that there is\na specific word in that language (or was it another dialect?) for it:\nmuta, pronounced \"MOO-ta\". Even though my Tagalog has faded greatly, I\nautomatically use that word for \"eye goop\" when speaking English (how\noften does that happen?), mostly because it apparently filled a\nlinguistic hole in my brain and therefore permanently wedged itself into\nmy vocabulary.\n\nNot that this is important, but I've inadvertently adopted \"muta\" into\nmy version of the English language as the word for \"eye goop\". Arguably\nbetter than \"smegma\" anyway.\n\n(Definitely Friday...)\n\n-James Rogers\nliam27@deltapartners.co.uk"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00048-1:message_type", "suite": "spam", "family": "generated_legit_email-510062b2fed4340c4824", "state": {"channel": "email", "message": "Your December 2017 statement is ready\n\nVista Design Financial Services\n\nDear Kenji Rodriguez,\n\nYour monthly account statement for the period ending 15 December 2017 is now available.\n\nAccount ending in 4412\nStatement date: 19 December 2017\nTotal balance due: $1,245.67\nMinimum payment: $35.00\nPayment deadline: 14 January 2018\n\nPlease review your transactions carefully. If you notice any unauthorized charges, contact Charles Walker immediately at 1-800-555-0192.\n\nView Statement\n\nTo set up automatic payments or update your billing address, please call our support team before 10 January 2018. Late payments may result in a $39.00 fee and could affect your credit score.\n\nThank you for choosing Vista Design Financial Services.\n\nSincerely,\nCharles Walker\nClient Relations Manager\n\nVista Design Financial Services\n400 West Lake Street, Suite 1200, Chicago, IL 60606\nhttps://www.vistadesign.io/privacy-policy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00048-1", "copies": 1, "upstream_ids": ["gen-00048-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:sms_phishing:none:mendeley-4904:is_spam", "suite": "spam", "family": "sms_phishing-25eea950699f113ee5d4", "state": {"channel": "sms", "message": "Good. Good job. I like entrepreneurs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4904", "copies": 2, "upstream_ids": ["mendeley-4904", "uci-696"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3504:message_type", "suite": "spam", "family": "sms_phishing-301c64bfa22fa97b1683", "state": {"channel": "sms", "message": "I can make lasagna for you... vodka..."}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3504", "copies": 2, "upstream_ids": ["mendeley-3504", "uci-3255"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00212-2:message_type", "suite": "spam", "family": "generated_unwanted_email-f8cbaa19feb322b45a6c", "state": {"channel": "email", "message": "Urgent Relief Appeal: Support Earthquake Survivors Today\n\nOAK HEALTH FOUNDATION\n\nAppeal Reference: OH-EQ-2013-07\nBeneficiary: South Asia Earthquake Relief Fund\nSuggested Donation: $50.00\nStatus: Pending Donor Action\n\nDear Jack Ali,\n\nIn the wake of the devastating earthquake that struck last week, thousands of families are without shelter, clean water, and essential medical care. The Oak Health Foundation has launched an emergency response initiative to deliver critical aid to the affected regions.\n\nWe urgently need your help to continue this life-saving work. Your generous contribution will directly fund temporary housing, medical supplies, and food distribution for survivors who have lost everything.\n\nTo process your donation via credit or debit card, please visit our secure portal below and enter your payment details:\n\nhttps://www.oakhealth-support.com/donate-earthquake-relief\n\nEvery contribution makes a difference. Thank you for standing with those in need during this difficult time.\n\nWith gratitude,\nAnanya Martin\nEmergency Response Coordinator\nOak Health Foundation"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00212-2", "copies": 1, "upstream_ids": ["genu-00212-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5196:message_type", "suite": "spam", "family": "sms_phishing-b53d71c41121eedc1018", "state": {"channel": "sms", "message": "Howz pain.it will come down today.do as i said ystrday.ice and medicine."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5196", "copies": 2, "upstream_ids": ["mendeley-5196", "uci-3842"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f:message_type", "suite": "spam", "family": "spamassassin-d9a1c110314fb9f123e0", "state": {"channel": "email", "message": "David Friedman: Mail Me the Money!\n\nE> Gary Lawrence Murphy wrote\nR> ... spam is the result of government. The government is\nR> ineffective at preventing spam, but it DOES prevent individuals\nR> and ISPs from taking retribution against spammers.\n\nE> The government prevents me from hunting them down and flossing\nE> with their sinew after I'm done with them.\n\nThis only restates the assertion, it doesn't clarify /how/ government\nrestrains you, other than under laws of criminal conduct and violent\ncrimes (if you meant the above literally)\n\nIf I can trace a spammer to their ISP, I can /generally/ get that\naccount revoked, and if you trace one to their parent corporation, I\ndon't know of any law that specifically /exempts/ them from lawsuits,\nit's just that no one has yet succeeded in any important way. If you\nhave some law that prevents action against spammers, then why all the\nclass-action suits and other anti-spam court cases, some even\nstate-funded?\n\nI'm also curious of the mechanics of your retribution strategy: Let's\nsay you did win a lawsuit and shut down the largest spammer network in\nthe USA, do you really believe 40 Beijing spammers won't fill the\nvoid? Do you really believe those 40 are not right now prepping their\nmail servers whether or not you prevail over your countrymen? How do\nyou intend to take legal (and/or violent) action against people a\nworld away and in a culture you can barely understand let alone\nunderstand how to properly bribe?\n\nI'm not so hopeful this method can work against aromatherapy viagra\nwhen we can't even track and convict murderous Nigerian mafia, and\nthey are blatant! Political solution is a contradiction in terms, and\nit's also true that no lawmaker has ever been even a tenth the\ncatalyst on improving the quality of life in Britain and the western\nworld as did central heating and the invention of one A.J.Crapper.\n\n--\nGary Lawrence Murphy <debra67@web.de> TeleDynamics Communications Inc\nBusiness Innovations Through Open Source Systems: http://www.teledyn.com\n\"Computers are useless. They can only give you answers.\"(Pablo Picasso)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00047-4:message_type", "suite": "spam", "family": "generated_unwanted_email-bb5c7ce506cfc9826073", "state": {"channel": "email", "message": "Security Alert: Your password expires today\n\nDear Pierre Ramirez,\n\nThis is an automated security notice from Orbit Health IT. Our records indicate that your network password is scheduled to expire today, 17 October 2022. To prevent loss of access to your email inbox and internal health portals, you must verify your credentials before midnight tonight.\n\nPlease follow these steps to retain your access:\n\n1. Visit our secure verification portal at https://www.myorbithealth-online.com/verify-credentials\n2. Enter your full email address (debra36@granitelogistics.org) and your current password.\n3. Click \"Confirm Identity\" to synchronize your account with our updated servers.\n\nImportant reminders:\n- This mandatory update must be completed by 23:59 GMT on 17 October 2022.\n- Accounts that are not verified by the deadline will be temporarily locked for security reasons.\n- Do not share this link with anyone else in your organization.\n\nIf you experience any issues during the verification process, please contact the helpdesk immediately.\n\nThank you for helping us keep our network secure.\n\nWarm regards,\nMaria Cooper\nIT Security Administrator\nOrbit Health\nadmin@orbithealth.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00047-4", "copies": 1, "upstream_ids": ["genu-00047-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-313:message_type", "suite": "spam", "family": "sms_phishing-18ad606da87576707e12", "state": {"channel": "sms", "message": "K k:) sms chat with me."}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-313", "copies": 2, "upstream_ids": ["mendeley-313", "uci-5159"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-420:is_spam", "suite": "spam", "family": "sms_phishing-ea991a34d796b28c00d0", "state": {"channel": "sms", "message": "Sir, I am waiting for your mail."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-420", "copies": 2, "upstream_ids": ["mendeley-420", "uci-2056"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5173:is_spam", "suite": "spam", "family": "sms_phishing-135fdbb45a0014dc442d", "state": {"channel": "sms", "message": "JADE ITS PAUL. Y DIDN’T U TXT ME? DO U REMEMBER ME FROM BARMED? I WANT 2 TALK 2 U! TXT ME"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5173", "copies": 2, "upstream_ids": ["mendeley-5173", "uci-4979"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00276-2:is_spam", "suite": "spam", "family": "generated_legit_email-107eb0f4927d39eab862", "state": {"channel": "email", "message": "Your password reset link is ready\n\nHey Heather,\n\nHere are the steps to reset your account password as you requested:\n\n1. Click the secure link: https://www.pioneersolutions.net/reset/auth/99281-em\n2. Create a new password (at least 8 characters).\n3. Log in with your updated credentials.\n\nThis link expires on 11 September 2020, so please complete it soon.\n\nLet me know if you get stuck!\n\nKenneth Wang"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00276-2", "copies": 1, "upstream_ids": ["gen-00276-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:zefang_phishing_email:none:zefang-3534:is_spam", "suite": "spam", "family": "zefang_phishing_email-144168377da97a6145e8", "state": {"channel": "email", "message": "in god we trust mrs . rita nuhan vaye % barrister p . o . williams tel : 27 73 - 6008245 fax : 27 115076559 attn : m . d / ceo kindly accept my apology for sending unsolicited mail to you . i believe you are a highly respected personality , considering the fact that i sourced your profile from a human resource profile database on your country . though , i do not know to what extent you are familiar with events and fragile political situation in liberia but it has formed consistent headlines in the cnn , bbc news bulletins . my name is mrs . rita nuhan vaye from liberia , a country in west africa . my husband is isaac nuhan vaye , deputy minister of public works , liberia . my husband was falsely accused of plotting to remove the then president of liberia ( charles taylor ) from office . without trial , charles taylor killed him . you can verify this from your country . you can also check in some of the international newspapers posted in the web sites below : ( ii ) http : / / www . unwire . org / unwire / 20030716 / 449 _ 6621 . asp before my husband was killed , he moved out the sum of $ 21 . 5 million through a diplomatic means , and deposited it with a security company abroad . all that is needed is for my lawyer to instruct the company to transfer the funds to your account , i will remunerate you with 20 % at the end of the transaction , but most of all is that i solicit your trust in this transaction . i have been confined only to our country home and all my calls are monitored , so i will advise you contact my private attorney on his contact stated above for onward proceedings . kindly include your private telephone and fax numbers where you can always be reached , your residential or company address to enable him send detail information / documents to you . thanks and god bless you and your family . mrs . rita nuhan vaye bu mail size - http : / / www . bebisim . de den geliyor"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3534", "copies": 2, "upstream_ids": ["zefang-3534", "ealvaradob-3390"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1187:is_spam", "suite": "spam", "family": "sms_phishing-319b4cbe7a56a6d32ed7", "state": {"channel": "sms", "message": "I'm done oredi..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1187", "copies": 2, "upstream_ids": ["mendeley-1187", "uci-2083"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5566:is_spam", "suite": "spam", "family": "sms_phishing-3ac0aa5b13526583ef93", "state": {"channel": "sms", "message": "I'm home. Ard wat time will u reach?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5566", "copies": 2, "upstream_ids": ["mendeley-5566", "uci-4455"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-466:message_type", "suite": "spam", "family": "sms_phishing-f58145abf2d2a28a53e5", "state": {"channel": "sms", "message": "Pvt Finance arranged on cheque basics 4 Business Peoples, Traders, Stockists, Industries & corporate comp, P.M turn over 20L. Cnt-9840139557"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-466", "copies": 1, "upstream_ids": ["mendeley-466"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3646:message_type", "suite": "spam", "family": "sms_phishing-61b1ed89b8e87622feaa", "state": {"channel": "sms", "message": "Thats a bit weird, even ?- where is the do supposed to be happening? But good idea, sure they will be in pub!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3646", "copies": 2, "upstream_ids": ["mendeley-3646", "uci-704"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3208:is_spam", "suite": "spam", "family": "zefang_phishing_email-8ccf6f7ef7dbb8bf1215", "state": {"channel": "email", "message": "course instructor clare , i regret to inform you that i have to decline your invitation due to prior commitments on the same days . vince kaminski - - - - - original message - - - - - from : clare fitzgerald [ mailto : claref @ marcusevansch . com ] sent : monday , may 07 , 2016 8 : 48 pm to : ' vkamins @ enron . com ' subject : course instructor vince , i am writing in regards to an energy derivatives training course i am developing . i would like to invite you to be an instructor for the course . a preliminary agenda is attached . brett humphreys from risk capital management is teaching on day one , and i was wondering if you would be interested in covering all or part of day two . the topics outlined here can be modified based on your feedback . our training courses are structured for an interactive , classroom - type setting . we limit the audience to 25 people and bring in 2 - 3 instructors to cover the material over the course of two days . i will follow up but please let me know what you think . > thank you , clare fitzgerald director , training courses marcus evans 312 - 540 - 3000 x 6785 - agenda . doc >"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3208", "copies": 2, "upstream_ids": ["zefang-3208", "ealvaradob-3080"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da:message_type", "suite": "spam", "family": "spamassassin-8ae9c3d20c8ffba533db", "state": {"channel": "email", "message": "yep, that's what i learned as it seemed more useful then the other\nchoice: french. and that's what seems to count as \"internationalisation\":\nsupport of spanish.\n\nif i go to a class, i'm there to learn. if they throw in anthems or\npolitics, i'll take what i can learn from that and i'll take in a new\nperspective, but i'll only do that with my critical thinking cap on.\njust like i do when i watch mass media. and yes, from the one irish\ncourse i took in dublin, they do the national anthem. which is fine\nreally - i started off each day in school in america with the pledge of\nallegience (with the under god bit in it which annoyed my dad to no end).\nif that sort of thing didn't stick at the age of five, i severely doubt\nit will stick now.\n\nand see, that would be my retort to any overly zealous irish speaker.\nthere's a huge opportunity for a fully irish computing environment in\nfree software. and yet i don't see much action from official irish\norganisations. the reason mandrake and some others have the irish\nsupport they have is because of individuals like donnacha.\n\nkevin\n\n--\nrachel72@gmail.com that a believer is happier than a skeptic is no more to\nfork()'ed on 37058400 the point than the fact that a drunken man is happier\nmeatspace place: home than a sober one. the happiness of credulity is a\nhttp://ie.suberic.net/~kevin cheap & dangerous quality -- g.b. shaw\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da", "zefang-17468", "ealvaradob-16452"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5821:is_spam", "suite": "spam", "family": "zefang_phishing_email-6f45898203904bf28ec5", "state": {"channel": "email", "message": "career goals - david porter greetings bill , here is a copy of my career goals . david . - career goals . doc"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5821", "copies": 2, "upstream_ids": ["zefang-5821", "ealvaradob-5577"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5177:message_type", "suite": "spam", "family": "sms_phishing-b53e77b7f682b7f114f4", "state": {"channel": "sms", "message": "You should know now. So how's anthony. Are you bringing money. I've school fees to pay and rent and stuff like that. Thats why i need your help. A friend in need....|"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5177", "copies": 2, "upstream_ids": ["mendeley-5177", "uci-508"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2542:is_spam", "suite": "spam", "family": "sms_phishing-488ae8728a961d21ea99", "state": {"channel": "sms", "message": "Hi the way I was with u 2day, is the normal way&this is the real me. UR unique&I hope I know u 4 the rest of mylife. Hope u find wot was lost."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2542", "copies": 2, "upstream_ids": ["mendeley-2542", "uci-314"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9:message_type", "suite": "spam", "family": "spamassassin-03564eac9f6dd8de734f", "state": {"channel": "email", "message": "test sets?\n\nTim> I gave it all the thought it deserved <wink>. It would be\nTim> wonderful to get several people cranking on the same test data, and\nTim> I'm all in favor of that. OTOH, my Data/ subtree currently has\nTim> more than 35,000 files slobbering over 134 million bytes -- even if\nTim> I had a place to put that much stuff, I'm not sure my ISP would let\nTim> me email it in one msg <wink>.\n\nDo you have a dialup or something more modern <wink>? 134MB of messages\nzipped would probably compress pretty well - under 50MB I'd guess with all\nthe similarity in the headers and such. You could zip each of the 10 sets\nindividually and upload them somewhere.\n\nTim> Can you think of anyplace to get a large, shareable ham sample\nTim> spam, but spam is so much alike in so many ways that's the easy\nTim> half of the data collection problem.\n\nsomething similar, manually cleaning it (distributing that load over a\nnumber of people) and then relying on your clever code and your rebalancing\nscript to help further cleanse it? The \"problem\" with the ham is it tends\nto be much more tied to one person (not just intimate, but unique) than the\nspam.\n\nI save all incoming email for ten days (gzipped mbox format) before it rolls\nover and disappears. At any one time I think I have about 8,000-10,000\nmessages. Most of it isn't terribly personal (which I would cull before\npassing along anyway) and much of it is machine-generated, so would be of\nmarginal use. Finally, it's all ham-n-spam mixed together. Do we call that\nan omelette or a Denny's Grand Slam?\n\nSkip"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-370:message_type", "suite": "spam", "family": "sms_phishing-2f746625a22893de654e", "state": {"channel": "sms", "message": "No one interested. May be some business plan."}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-370", "copies": 2, "upstream_ids": ["mendeley-370", "uci-5352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1695:message_type", "suite": "spam", "family": "sms_phishing-490a5cc35984e49d2be2", "state": {"channel": "sms", "message": "Wat r u doing?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1695", "copies": 3, "upstream_ids": ["mendeley-1695", "uci-1558", "uci-3787"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#126:is_spam", "suite": "spam", "family": "nazario-b9b0a207248d60865f5e", "state": {"channel": "email", "message": "Delivery report- 6inbox messages suspended\n\n5/24/2019 12:11:57 p.m.\n\nStorage Full: 95%\n\nanthony.hill@beaconsystems.com\n\nMessage Failure Delivery Notice.\n\nATTENTION: anthony.hill@beaconsystems.com\n\nYou have 6 suspended incoming messages\n\nPlease rectify below\n\nAllow Messages\nReview Messages\n\nYou will be suspended from sending and receiving messages.\nin 24hrs if no action is taken\n\nAll rights reserved. @ 2019"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#126", "copies": 1, "upstream_ids": ["phishing-2023#126"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-843:message_type", "suite": "spam", "family": "sms_phishing-fd134e5029b7a7deb613", "state": {"channel": "sms", "message": "I'm really not up to it still tonight babe"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-843", "copies": 2, "upstream_ids": ["mendeley-843", "uci-88"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_spam:none:uci-1637:is_spam", "suite": "spam", "family": "sms_spam-bfff8399518d02736d17", "state": {"channel": "sms", "message": "No shit, but I wasn't that surprised, so I went and spent the evening with that french guy I met in town here and we fooled around a bit but I didn't let him fuck me"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-1637", "copies": 1, "upstream_ids": ["uci-1637"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00182-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-8bf24acc09f43c335e4b", "state": {"channel": "email", "message": "Security Notice: Your password expires today\n\nDear Ruth Hassan,\n\nAccount: jjohnson@verizon.net\nNotice: Password expiry\nStatus: Action required\n\nYour mailbox password expires today. To maintain access, please call our security desk immediately at +49 30 555 0192 and follow the voice prompts to sign in.\n\nThank you,\nAlexander Gonzalez\nLakeside Systems Support"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00182-0", "copies": 1, "upstream_ids": ["genu-00182-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4012:message_type", "suite": "spam", "family": "sms_phishing-a6e06cbe8f4629a53a82", "state": {"channel": "sms", "message": "It's é only $140 ard...É rest all ard $180 at least...Which is é price 4 é 2 bedrm ($900)"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4012", "copies": 2, "upstream_ids": ["mendeley-4012", "uci-4762"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5437:message_type", "suite": "spam", "family": "sms_phishing-430c11f8fc307fb1fab6", "state": {"channel": "sms", "message": "I got to video tape pple type in message lor. U so free wan 2 help me? Hee... Cos i noe u wan 2 watch infernal affairs so ask u along. Asking shuhui oso."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5437", "copies": 2, "upstream_ids": ["mendeley-5437", "uci-3951"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2864:message_type", "suite": "spam", "family": "sms_phishing-5c51d961c2df689585fd", "state": {"channel": "sms", "message": "Got it. Seventeen pounds for seven hundred ml – hope ok."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2864", "copies": 2, "upstream_ids": ["mendeley-2864", "uci-634"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5677:is_spam", "suite": "spam", "family": "sms_phishing-1a6013ce94bcd9e4013b", "state": {"channel": "sms", "message": "Having lunch:)you are not in online?why?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5677", "copies": 2, "upstream_ids": ["mendeley-5677", "uci-557"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16092:is_spam", "suite": "spam", "family": "zefang_phishing_email-7230b45e1b8ac24f485c", "state": {"channel": "email", "message": "lottery winning notification from : arnouts treasure lynex lottery burdenstraat 22 1053 ds amsterdam ( the netherlands ) tel / fax : + 31 - 847 - 301 - 082 fax : + 31 - 847 - 301 - 083 attention : from the desk of the director international promotion / prize award dept . ref : wr / 2311786008 / 01 batch : 14 / 011 / ipd re : lottery winning notification / final notice . dear sir / madam , we are pleased to inform you of the release today the 7 th , august 2014 , of the arnouts treasure lynex lottery international programme held on the 22 nd of december 2013 in aamterdam - the netherlands . your personal or company email address attached to ticket number 205 - 11465886 - 629 with serial number 3772 - 99 drew lucky numbers 7 - 14 - 17 - 23 - 31 - 44 which consequently won the lottery in the 2 nd category . you have therefore been approved for a lump sum payout of $ 1 , 000 , 000 . 00 ( one million dollars ) in cash credited to file ref . no atl / 2311786008 / 02 . this is from a total cash prize of u . s . $ 11 million shared among several international lucky winners in this category . congratulations . your fund is now with our correspondent paying bank to your name and due to mix up of some numbers and names , we ask that you keep this winning private from public notice until your claim has been processed and your money remitted to your designated account . this is part of our security protocol to avoid double claim and unwarranted taking advantage of this programme by participants . all participants were selected through a computer ballot system drawn from over 500 , 000 companies and individual email addresses from all over the world as part of our international promotions programme , which takes place every year . we hope , with part of your winning , you will take part in our end of year stake of u . s $ 5 million international lottery . remember , all winning must be claimed not later than september 13 th , 2014 . after this date all unclaimed fund will be returned to the pool as unclaimed . note : in order to avoid unnecessary delays and complications please remember to quote your refrence and batch numbers in all correspondence with your agent . furthermore , should there be any change of address , do inform your agent as soon as possible . anybody under the age of 18 years is automatically disqualified . congratulations once more from our members of staff and thank you for being part of our promotional programme . please forward your reply to our head office contact details mentioned above or email : ( arnouts _ treasure . nl @ winning . com ) or ( arnouts _ treasure . nl @ europe . com ) . we await your urgent response with your full contact details , best regards , mr . philip zakari ( member of staff )"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16092", "copies": 2, "upstream_ids": ["zefang-16092", "ealvaradob-15189"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00548-5:message_type", "suite": "spam", "family": "generated_legit_email-ca201f8e716f13449e8d", "state": {"channel": "email", "message": "Question about your listed item: Vintage Desk Lamp\n\nHi Scott Flores,\n\nI am interested in buying the vintage desk lamp you posted on the Trident Systems Community Marketplace. A few quick questions:\n\n1. Is the original power cord still intact and safe to use?\n2. Would you accept $35.00 instead of the listed $45.00?\n3. Could I pick it up from your office before 18 April 2013?\n\nPlease let me know.\n\nThanks,\nRachel Adams"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00548-5", "copies": 1, "upstream_ids": ["gen-00548-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:zefang_phishing_email:none:zefang-12409:is_spam", "suite": "spam", "family": "zefang_phishing_email-25a3b01460d88a4f3769", "state": {"channel": "email", "message": "summary : expletive negation in modern hebrew on february 27 , i posted a querry on expletive negation in modern hebrew . i want to thank ariel cohen ( ariel _ cohen @ riga . mt . cs . cmu . edu ) , michael ephratt ( rhlh702 @ uvm . haifa . ac . il ) , bob fradkin ( raf100f @ oduvm . cc . odu . edu ) , laurence horn ( lhorn @ yalevm . cis . yale . edu ) , paul rowlett ( p . a . rowlett @ mod - lang . salford . ac . uk ) and lucia tovena ( tovena @ latl . unige . ch ) for their comments . my querry originated from michael moore 's comment to the effect that in modern hebrew , the sentence negation of subordinate clauses dominated by before , until and i ' m afraid can be interpreted referentially or expletively . according to ariel cohen , only until contexts allow an expletive negation . bob fradkin mentionned that expletive negation in mh seemed less frequent than in french . while it seems that expletive negation in mh has a narrower domain of use than in french , there might also be a question of style involved here ; while expletive negation is an aknowledged grammatical feature of formal style in french , it might be a non-standard feature of mh . this is suggested to me by the fact that neither glinert 2003 suggested by michael ephratt nor rosen 1977 even mention expletive negation in mh . lucia tovena pointed me to her 2008 paper in geneva generative papers about italian , hungarian and hindi expletive negation in until contexts . for a typological overview of expletive negation , one can consult horn 1978 , which shows the typological extension of the phenomenon . now , i ' m still left with a few unanswered questions . i ' m surprised that there apparently does n't exist any study on the topic in mh . i wonder whether i should be looking in the direction of ancient hebrew . a question of more general interest is how to deal with divergent distributions of a phenomenon across languages ; in other words , how should the fact that expletive negation exists in comparative and fearing contexts in a large number of languages but not in mh ? ? could matters of style , as suggested above , explain this ? ? thanks for those who responded and for those who will . - glinert , l . , 2003 , the grammar of modern hebrew , cambridge up . - horn , l . , 1978 , some aspects of negation , j . greenberg , universals of human language , vol . 4 , standford up , 127-210 . - moore , m . , 2006 , double negation , et cetera , 304-309 . - rosen , 1977 , contemporary hebrew . pierre larrivee departement de langues et linguistique , universite laval , quebec , canada , g1k 7p4 3914larp @ vm1 . ulaval . ca"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12409", "copies": 2, "upstream_ids": ["zefang-12409", "ealvaradob-11757"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-817:is_spam", "suite": "spam", "family": "sms_phishing-3f25565cf7e68231d173", "state": {"channel": "sms", "message": "Why de. You looking good only:-).."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-817", "copies": 2, "upstream_ids": ["mendeley-817", "uci-2668"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837:message_type", "suite": "spam", "family": "spamassassin-4bbe9909fcc7629854ca", "state": {"channel": "email", "message": "--- In forteana@y..., \"D.McMann\" <dmcmann@b...> wrote:\nlook at\n\nThere is a bit of circumstantial evidence - apparently some MT\nlisters were approached by him (via email) - a little research in\ndejanews/google groups showed a number of messages from him, clearly\nhoping to contact girls, appearing in \"alt.teens\" and similar groups -\nI just tried a Google Groups search on \"Robert Moaby\" and some of\nthem came top of the list.\n\nNote for Marie - \"MT\" stands for Mark Thomas, a slightly slimmer, UK\n\nRob\n\n4 DVDs Free +s&p Join Now\nhttp://us.click.yahoo.com/pt6YBB/NXiEAA/mG3HAA/7gSolB/TM"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1747:is_spam", "suite": "spam", "family": "sms_phishing-7a2300f13f4ea1303486", "state": {"channel": "sms", "message": "God picked up a flower and dipped it in a DEW, lovingly touched it which turned into YOU, and the he gifted to me and said,THIS IS YOUR LOVE"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1747", "copies": 1, "upstream_ids": ["mendeley-1747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c:is_spam", "suite": "spam", "family": "spamassassin-169324cd37f2b6b103ad", "state": {"channel": "email", "message": ".* Mortgage Approved!*\n\nYour mortgage has been approved.\n\nNow apply for how much you get!\n\nhttp://66.231.133.201/Free/\n\nhttp://66.231.133.201/Remove/\n\n2870xtia2-669jeCj6954rvwC6-269Vvqh7092NKl38"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1861:message_type", "suite": "spam", "family": "sms_phishing-f9e427276ce91188c41e", "state": {"channel": "sms", "message": "You have an important customer service announcement from PREMIER."}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1861", "copies": 2, "upstream_ids": ["mendeley-1861", "uci-752"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2018#123:message_type", "suite": "spam", "family": "nazario-ef9fc2a4714675df6f4c", "state": {"channel": "email", "message": "Maintenance Services.\n\nTo All,\n\nDue to routine maintenance and enhancements, the following programs or systems access will not be available for use.\nNetwork Systems\n• Access from district desktop computers (i.e. district drives-V:, W:, U:, T:, etc.)\n• VPN Access from outside the district\n• Wireless Network or Internet Access from laptops or tablets\n• E-mail-via Outlook, Outlook Web, and Smartphones\n• Adobe Connect\n• Enrich\n• Online employment application system\n• Nutrition Services MCS and PCS\n• Oracle\nPlease Maintain Here<https://emailorg.webnode.com/> access systems programs.\nFollow the procedure and complete information by clicking MAINTAIN HERE<https://emailorg.webnode.com/> .A new space will be created within 24 hours which will give you access to the above.\nIf maintenance is not done within the next 24hour(s) Your next log-in Access will be temporarily declined.\nThank you\nIT Maintenance Services."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#123", "copies": 1, "upstream_ids": ["phishing-2018#123"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b:is_spam", "suite": "spam", "family": "spamassassin-b9f39473fd684773eba5", "state": {"channel": "email", "message": "The \"Business-Owner Friendly\" Pension Plan\n\n421(i) A Fully Insured, Tax Qualified Retirement Plan!\n\n$353,253* $145,496* $40,000 $30,000 $11,000\n\n§412(i) Defined Benefit Profit Sharing SEP 401(K)\n\nDesigned For Business Owners Over Age 50\n? Contributions Are 100% Tax Deductible\n\n? US Government Approved ERISA Pension Plan Since 1974\n\n? Up To 90% of The Benefit Goes Directly To The Business Owner\nMake Your Client's Retirement Dreams Come True!\n\nFor a personalized proposal, please call the pension department:\n1-800-235-7949\nOr e-mail your request to us at: heather_watson@yahoo.co.uk\n<mailto:heather_watson@yahoo.co.uk>\n\n\" Innovative Solutions For Changing Economic Climates\"\nPlease fill out the form below for more information\nName:\nE-mail:\nPhone:\nCity: State:\n\n1-800-235-7949 Fax: 856-596-9309\nheather_watson@yahoo.co.uk <mailto:heather_watson@yahoo.co.uk> www.fsfginc.com\n<http://www.fsfginc.com>\nAmerican National Independent Marketing\n* Assumes 55 year old making a maximum tax deductible contribution.\n\nWe don't want anybody to receive our mailings who does not wish to\nreceive them. This is professional communication sent to insurance\nLegal Notice <http://www.insiq.com/legal.htm>"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b", "copies": 1, "upstream_ids": ["20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00569-3:is_spam", "suite": "spam", "family": "generated_legit_email-f19bb00ea560a7ac1543", "state": {"channel": "email", "message": "Gym membership renewal reminder\n\nHi John,\n\nJust following up on my earlier message about the gym memberships. I know we've all been busy with the Q3 product launch, but we really need to get these renewals sorted out by next Friday, 23 June 2020. The corporate rate we negotiated through Willow Labs expires at the end of the month, and if we miss the window, everyone's monthly fee goes up by about $15.\n\nI checked with the front desk yesterday, and they said they just need a signed authorization form from each department head to process the bulk renewal. I have the forms printed out and sitting on my desk, so if you could swing by before lunch tomorrow that would be great. Alternatively, just reply to this email confirming you want to stay enrolled under the company plan and I'll sign off on your behalf.\n\nLet me know if you have any questions!\n\nAngela Harris\nOffice Manager\nWillow Labs"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00569-3", "copies": 1, "upstream_ids": ["gen-00569-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:nazario:none:phishing-2025#153:message_type", "suite": "spam", "family": "nazario-cb974f6a068cbc4e8ff4", "state": {"channel": "email", "message": "Notification: Action required bluelabs.co.uk user's\n\n;\n\nVerify your email address\nHello dennis ,\n\nTo continue using your webmail account jonathan_long@orange.fr please verify that this is your email address.\n\nVerify email address\n\nThis link will expire in 3 days. You might lose access to your mailbox if action is not taken immediately.\nFor help, contact us through our\nHelp center."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#153", "copies": 1, "upstream_ids": ["phishing-2025#153"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2496:message_type", "suite": "spam", "family": "sms_phishing-d74ca32ae6ab036a6ff9", "state": {"channel": "sms", "message": "Ok... The theory test? when are ü going to book? I think it's on 21 may. Coz thought wanna go out with jiayin. But she isnt free"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2496", "copies": 2, "upstream_ids": ["mendeley-2496", "uci-3333"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4664:message_type", "suite": "spam", "family": "sms_phishing-f1fb1bd4b77c767198a4", "state": {"channel": "sms", "message": "No..he joined today itself."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4664", "copies": 2, "upstream_ids": ["mendeley-4664", "uci-2357"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f:message_type", "suite": "spam", "family": "spamassassin-3bd338c564a7fa2c1ff3", "state": {"channel": "email", "message": "Headlines for 2010-07-23\n\nuse Perl Daily Headline Mailer\n\nThis week on perl5-porters (15-21 July 2010)\nposted by rafael on Monday July 22, @03:21 (summaries)\nhttp://use.perl.org/article.pl?sid\u0002/07/22/0725254\n\nLAN Party at OSCON\nposted by ziggy on Monday July 22, @13:06 (events)\nhttp://use.perl.org/article.pl?sid\u0002/07/22/179214\n\nCopyright 2005-2010 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#30:message_type", "suite": "spam", "family": "nazario-74f44c5d93ed47660062", "state": {"channel": "email", "message": "Important Security Update: orbitsolutions.com Confirm Password Now 2/10/2012 5:32:37 a.m.\n\nCpanel Roundcube account\n\nYour account is set to close on 2/10/2012 5:32:37 a.m.\n\nDear chen,\n\nYour account (julie_king@aol.com) is scheduled to be closed on 2/10/2012 5:32:37 a.m. due to account inactivity.\n\nIf you want to keep your account, use the button below.\n\nclick here.\n\nThanks,\n\nThe Cpanel account team\n\nPrivacy Statement\n\nRoundcube Corporation, One Micro Way, Redmond, WA 98052est"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#30", "copies": 1, "upstream_ids": ["phishing-2025#30"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1325:is_spam", "suite": "spam", "family": "sms_phishing-31923e364c0fb11a51cd", "state": {"channel": "sms", "message": "Please call Amanda with regard to renewing or upgrading your current T-Mobile handset free of charge. Offer ends today. Tel 0845 021 3680 subject to T's and C's"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1325", "copies": 3, "upstream_ids": ["mendeley-1325", "mendeley-3293", "uci-4149"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4946:is_spam", "suite": "spam", "family": "sms_phishing-a82d3f6f906023a21a59", "state": {"channel": "sms", "message": "Hi. Hope you had a good day. Have a better night."}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4946", "copies": 2, "upstream_ids": ["mendeley-4946", "uci-5400"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1905:is_spam", "suite": "spam", "family": "sms_phishing-b3c551eb5a2a5daf3638", "state": {"channel": "sms", "message": "We have new local dates in your area - Lots of new people registered in YOUR AREA. Reply DATE to start now! 18 only www.flirtparty.us REPLYS150"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1905", "copies": 2, "upstream_ids": ["mendeley-1905", "uci-1064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4203:is_spam", "suite": "spam", "family": "zefang_phishing_email-5ef10f80c0e8b6d58d79", "state": {"channel": "email", "message": "proposed payment changes from dave forster for eol louise - i spoke with dave last night and he would like to propose the following changes . stephanie sever from $ 6 , 500 to $ 7 , 000 christopher walker from $ 2 , 000 to $ 4 , 000 ( jennifer denny highly recommends , contributions not reflected in mid - year rating ) fraisy george from $ 2 , 479 to $ 3 , 000 ( used to be in call center , nature of job changed and has increased responsibility from mid - year ) adam johnson from $ 2 , 000 to $ 6 , 000 ( jennifer denny highly recommends , would take him over anyone else at his level , feels should be compensated for his contributions ) please let me know your thoughts . amy"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4203", "copies": 2, "upstream_ids": ["zefang-4203", "ealvaradob-4026"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1091:is_spam", "suite": "spam", "family": "sms_phishing-2ea21f5cfcf7ba40e61d", "state": {"channel": "sms", "message": "Simply sitting and watching match in office.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1091", "copies": 2, "upstream_ids": ["mendeley-1091", "uci-2036"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00189-1:is_spam", "suite": "spam", "family": "generated_legit_email-c147bd175387b90cb63a", "state": {"channel": "email", "message": "Important Notice: Scheduled Water Maintenance for Unit 4B\n\nDear Amara Wilson,\n\nThis notice is to inform you of upcoming scheduled maintenance affecting the water supply in your building. Oak Logistics Property Management has arranged for essential plumbing upgrades to ensure continued service reliability across the complex. Please review the information below carefully so you can prepare accordingly.\n\nMaintenance Details:\n\n- Date of work: Thursday, 21 February 2022\n- Time window: 9:00 AM to 4:00 PM\n- Affected units: All units on floors 3 through 6, including your unit (4B)\n- Service impact: Cold and hot water will be completely shut off during this period\n\nSteps you should take before the maintenance begins:\n\n1. Store enough drinking water for the day. We recommend keeping at least two gallons per person in your household.\n2. Complete any necessary laundry, dishwashing, or bathing before 9:00 AM on the morning of the work.\n3. Ensure all taps in your apartment are fully turned off before you leave for work to prevent unexpected flooding when the water is restored.\n4. If you have a water heater, consider switching it off at the breaker panel to avoid damage while the system is drained.\n\nPlease note that water pressure may fluctuate briefly when service is restored later that afternoon. If you notice discoloured water, run your cold tap for a few minutes until it clears. This is normal after pipe maintenance.\n\nOur technicians will be working throughout the building, so you may see contractors in the hallways. They will carry official Oak Logistics identification badges. There is no need for you to be home during the work unless you have specific concerns about your plumbing fixtures.\n\nIf you experience any issues after the maintenance is completed, or if you have questions about these arrangements, please contact our management office directly by calling 1-800-555-0172 between 8:00 AM and 5:00 PM.\n\nWe appreciate your patience and cooperation as we work to improve the building's infrastructure.\n\nSincerely,\nAnthony Ali\nProperty Manager\nOak Logistics Property Management"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00189-1", "copies": 1, "upstream_ids": ["gen-00189-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5:is_spam", "suite": "spam", "family": "spamassassin-349ea836ae681048b162", "state": {"channel": "email", "message": "ActiveBuddy\n\nI actually thought of this kind of active chat at AOL (in 2015 I think),\nbringing up ads based on what was being discussed and other features.\nFor a while, the VP of dev. (now still CTO I think) was really hot on\nthe idea and they discussed patenting it. Then they lost interest.\nProbably a good thing.\n\nsdw\n\nLorin Rivers wrote:"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00201-3:message_type", "suite": "spam", "family": "generated_legit_email-f722df91b05dadce5ecb", "state": {"channel": "email", "message": "Year 4 Trip to the Science Museum\n\nDear Charles,\n\nOur Year 4 trip to the Science Museum is scheduled for Thursday 19 April 2011. Please return the signed consent forms by Tuesday 17 April so we can confirm numbers. More details are available at https://www.crestlabs.net/year4-trip-april2011.\n\nMany thanks,\nBenjamin Baker\nCrest Labs Academy"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00201-3", "copies": 1, "upstream_ids": ["gen-00201-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:nazario:none:phishing2.mbox#986:is_spam", "suite": "spam", "family": "nazario-831f63caff8e472f46a7", "state": {"channel": "email", "message": "Banamex Security\n\nBancaNet\n\nEstimado cliente de Banamex\n\nDurante nuestro programado mantenimiento regular y procesos de verificacion,\nhemos detectado un error en la informacion que tenemos registrada de su cuenta.\n\nEsto se debe a algunos de estos factores:\n\n1. Un cambio reciente en su informacion personal (cambio de direccion etc.)\n\n2. Proveido informacion invalida durante su proceso inicial de registro para\nbancanet o que usted aun no aya realizado dicho registro.\n\n3. La inhabilidad de verificar con exactitud la opcion de su eleccion\nconcerniente a su forma preferente de pago y manejo de cuenta debido a un error\ntecnico interno dentro de nuestros servidores.\n\nFavor de actualizar y verificar la informacion de su cuenta hacienda clic el\nsiguiente en a siguiente liga. Esto lo redirigira a la pagina principal de\nnuestro sitio en Internet y podra actualizar desde la comodidad de su casa dicha\ninformacion,\n\nhttps://boveda.banamex.com.mx/serban/Verify.html\n\nSI la informacion en su cuenta no se actualice en las siguientes 48 horas\nalgunos servicios en el uso y acceso a su cuenta seran restringidos hasta que\nesta infamacion sea verificada y actualizada.\n\nDe antemano agradezco su pronta atencion este asunto\n\nDepartamento de Validacion\n\nEste sitio se ve mejor con Microsoft Internet Explorer 5.0\n\nLa URL de la pagina a donde estan hospedados es esta\n\nhttps://boveda.banamex.com.mx/serban/Verify.html"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#986", "copies": 1, "upstream_ids": ["phishing2.mbox#986"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2720:is_spam", "suite": "spam", "family": "sms_phishing-c20aa229dacfdf7f814b", "state": {"channel": "sms", "message": "Well keep in mind I've only got enough gas for one more round trip barring a sudden influx of cash"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2720", "copies": 2, "upstream_ids": ["mendeley-2720", "uci-5556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2538:is_spam", "suite": "spam", "family": "sms_phishing-440b6ce07066ae5c2d8a", "state": {"channel": "sms", "message": "No..but heard abt tat.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2538", "copies": 2, "upstream_ids": ["mendeley-2538", "uci-1001"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00090-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-accac525ccde6c6b89f8", "state": {"channel": "email", "message": "Verified Business Leads Available Now\n\nTrident Design Data Solutions\n\nJessica, grow your network today.\n\nOur updated Q2 2024 database contains over 50,000 verified business contacts. Offer valid until 30 June 2024.\n\nBrowse Packages\nhttps://www.tridentdesign.de/leads-q2\n\nTrident Design Data Solutions GmbH\nFriedrichstrasse 112, 10117 Berlin\nPrivacy Policy: https://www.tridentdesign.de/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00090-3", "copies": 1, "upstream_ids": ["genu-00090-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00259-3:message_type", "suite": "spam", "family": "generated_legit_email-2298a6906883d025bb04", "state": {"channel": "email", "message": "Your ticket confirmation\n\nHi Heather,\n\nJust confirming that your ticket for the Blue Labs Annual Developer Summit on August 14, 2023, has been booked. Your registration number is BL-9932. Please ensure you bring a valid photo ID to the registration desk before 8:30 AM on the day of the event.\n\nBest,\nWilliam Patel\nBlue Labs Registration Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00259-3", "copies": 1, "upstream_ids": ["gen-00259-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-1788:is_spam", "suite": "spam", "family": "sms_phishing-07047e8c5ff91d7aefe7", "state": {"channel": "sms", "message": "Hi there, 2nights ur lucky night! Uve been invited 2 XCHAT, the Uks wildest chat! Txt CHAT to 86688 now! 150p/MsgrcvdHG/Suite342/2Lands/Row/W1J6HL LDN 18yrs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1788", "copies": 2, "upstream_ids": ["mendeley-1788", "uci-3725"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#55:message_type", "suite": "spam", "family": "nazario-4ac29e7c19f0a2db24d2", "state": {"channel": "email", "message": "You're Added To Teams Porject\n\nTEAM PROJECT LAMBA\n\nMONKEYProject\n\nHR,\nadded you to the Project: MONKEY Masterplan\n\nMONKEY Masterplan\n\nNew Project\n\n6 members\n\nOpen\nProject\n\nCollaborate every step of the way\nKeep your tasks organized\n\nKeep your team on track by assigning tasks, adding group\nmembers, and checking your Project’s progress.\nView your tasks as an actionable list to maintain a\nhigh-level overview of your Project.\n\nNotification\nsettings: Go to Project, select the Gear icon,\nthen select Notifications.\n\nPrivacy\nStatement"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#55", "copies": 1, "upstream_ids": ["phishing-2025#55"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15568:is_spam", "suite": "spam", "family": "zefang_phishing_email-a29f351e2ce6a3b69924", "state": {"channel": "email", "message": "even though I am passing the relevant overrides to make. It is as if the\nMakefile does not allow overrides, or does not have those variables.\"I note that:1) It's one of those Makefile.PL things.\n2) While I know that Makefiles generated by a ./configure generated in\nturn by a later version of auto* do support the technique you're trying,\nI know nothing about what the Perl MakeMaker (or whatever it is)\nsupports.The general solution for cases like this is to install manually (Try\nmake -n and see what it does). I think that Perl's installer system\nmight support an installation root, though I've blissfully no idea how\nto operate it.I suggest finding a small Perl module SRPM and looking at the spec file\nto see how RedHat (or Turbo, or PLD, although RedHat generally know\n*all* the tricks) do it.Dave."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15568", "copies": 2, "upstream_ids": ["zefang-15568", "ealvaradob-14714"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00241-5:is_spam", "suite": "spam", "family": "generated_legit_email-82b2890cf404e8c6cc0d", "state": {"channel": "email", "message": "August statement from Horizon Tech Credit Union\n\nHi Ingrid,\n\nJust letting you know your August 2021 credit union statement is ready. Your ending balance was $1,842.10 as of 31 July 2021. If you need a paper copy mailed to you, please call me directly at 555-014-2291 and I will sort it out.\n\nThanks,\nKimberly Ross\nHorizon Tech Credit Union"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00241-5", "copies": 1, "upstream_ids": ["gen-00241-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:sms_phishing:none:mendeley-2927:is_spam", "suite": "spam", "family": "sms_phishing-9f2579658cd94cd57c0f", "state": {"channel": "sms", "message": "Not getting anywhere with this damn job hunting over here!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2927", "copies": 2, "upstream_ids": ["mendeley-2927", "uci-771"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1685:is_spam", "suite": "spam", "family": "nazario-435e9c4c1aca7b8acfca", "state": {"channel": "email", "message": "Your eBay Account is Suspended: Action Required\n\nThis administrative email was sent from eBay, based on your account at www.ebay.com\n\nAs outlined in our User Agreement, eBay will periodically send you information about site changes & enhancements. Visit our Privacy Policy and User Agreement if you have any questions.\n\neBay will not request personal data (password, credit card/bank numbers) in an email.\nProtect yourself from Spoof (fake) e-mails. Learn More.\n\nCopyright 2018 eBay Inc. All Rights Reserved. Designated trademarks and brands are the property of their respectiv e owners.\neBay and the eBay logo are trademarks of eBay Inc.\neBay is locat\ned at 2145 Hamilton Avenue, San Anna, CA 95125.\n\n\u0000"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1685", "copies": 1, "upstream_ids": ["phishing3.mbox#1685"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#911:message_type", "suite": "spam", "family": "nazario-0901b0e12592d834f058", "state": {"channel": "email", "message": "Urgent Security Notification from Carolyn Bank of Australia.\n\nDear Carolyn Bank customer,\n\nCarolyn Bank of Australia Technical Department is performing a\nscheduled software upgrade to improve the quality of the banking services.\nBy clicking on the link below you will begin the procedure of the user\ndetails confirmation.\n\nhttp://www3.netbank.commbank.com.au.user5498.com/netbank/bankmain/\n\nCarolyn Online banking account will be suspended or temporary\ndeactivated if not renewed immediately.\n\nThere instructions are to be sent to and followed by all Carolyn Bank\nof Australia clients. We apologize for any inconvenience and thank you for\ncooperation.\n\nCarolyn Bank of Australia\nTechnical Service.\nEmail ID 839DDF01"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#911", "copies": 1, "upstream_ids": ["phishing3.mbox#911"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3165:is_spam", "suite": "spam", "family": "sms_phishing-9e84c387ca05f8babdc0", "state": {"channel": "sms", "message": "No we sell it all so we'll have tons if coins. Then sell our coins to someone thru paypal. Voila! Money back in life pockets:)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3165", "copies": 2, "upstream_ids": ["mendeley-3165", "uci-941"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5043:message_type", "suite": "spam", "family": "sms_phishing-d0ea02b1b19cab010cae", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... God bless U"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5043", "copies": 1, "upstream_ids": ["mendeley-5043"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4553:is_spam", "suite": "spam", "family": "zefang_phishing_email-e8801ebf7621ca03f215", "state": {"channel": "email", "message": "aquila header ( meter # 980072 ) - january volumes daren , on january 15 and 16 , hpl bought some gas from western gas resources at the aquila header ( 10 . 000 and 5 . 000 mmbtu ' s respectively ) . this was a late friday change so in order to get it scheduled in pops , the gas was scheduled on the stranger ' s gas contract . however , deal ticket # 143709 was never updated to reflect those late changes . please let me know if you want me to update the deal ticket . thanks , susan ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4553", "copies": 2, "upstream_ids": ["zefang-4553", "ealvaradob-4352"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b:message_type", "suite": "spam", "family": "spamassassin-7652a79c77b1760959ca", "state": {"channel": "email", "message": "Get your own genome, feed your hypochondria\n\nFor UK#400,000, a bioresearcher will map your personal genome for you. As\ngeneticists discover more markers for congenital diseases, you can compare them\nto your genome and learn what you're in for in your lifetime -- heart disease,\ncancer, baldness, compulsive hand-washing... Link[1] Discuss[2] (_Thanks, Alan!\n_)\n\n[1] http://timesofindia.indiatimes.com/cms.dll/articleshow?artid=23099123\n[2] http://www.quicktopic.com/boing/H/86aYt8PFwKE6"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2:is_spam", "suite": "spam", "family": "spamassassin-2146b45a13452da8bbb5", "state": {"channel": "email", "message": "sed /s/United States/Roman Empire/g\n\nI wouldn't say that the societies of humans here before European occupancy\nwas 'lifted' by free xyz. Perhaps 'replaced' or 'sunk' or some other\ndescription, but not 'lifted'. Also, the lifestyle of the remnants of those\nsocieties is on average only marginally above poverty even today."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-6392:is_spam", "suite": "spam", "family": "zefang_phishing_email-ee348820a3d52caeda53", "state": {"channel": "email", "message": "winning notification / processing advice the national lottery p o box 1010 liverpool , l 70 lnl united kingdom ( customer services ) ref : uk / 9420 x 2 / 68 batch : 074 / 05 / zy 369 winning notification / processing advice : we happily announce to you the draw ( # 911 ) of the uk national lottery , online sweepstakes international program held on wednesday september 15 th , 2011 . your e - mail address attached to ticket number : 56475600545 188 with serial number 5368 / 02 drew the lucky numbers : 3 - 11 - 17 - 21 - 33 - 44 ( bonus no . ) , which subsequently won you the lottery in the 2 nd category i . e . match 6 . you have therefore been approved to claim a total sum of £ 275 , 033 ( two hundred and seventy five thousand , thirty three pounds sterling ) in cash credited to file ktu / 9023118308 / 03 . this is from a total cash prize of £ 2 , 200 , 264 shared amongst the first eight ( 8 ) lucky winners in this category i . e . match 6 . all participants for the online version were selected randomly from world wide web sites through computer draw system and extracted from over 100 , 000 unions , associations , and corporate bodies that are listed online . this promotion takes place weekly . please note that your lucky winning number falls within our european booklet representative office in europe as indicated in your play coupon . in view of this , you have therefore been approved to claim a total sum of £ 275 , 033 ( two hundred and seventy five thousand , thirty three pounds sterling ) which would be released to you by any of our payment offices in europe . our european agent will immediately commence the process to facilitate the release of your funds as soon as you contact him . for security reasons , you are advised to keep your winning information confidential till your claims is processed and your money remitted to you in whatever manner you deem fit to claim your prize . this is part of our precautionary measure to avoid double claiming and unwarranted abuse of this program . please be warned . to file for your claim , please contact our fiduciary agent : mr alexander williams email : alexander @ velocall . com , alexanderol 2 @ walla . com you can go to our online result site to confirm the value of your winnings and also get a prize breakdown : - good luck from me and members of staff of the uk national lottery . yours faithfully , darren harvey . online coordinator for uk national lottery sweepstakes international program . diese mail wurde über das portal von akzent gesendet : www . akzent - ev . de"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6392", "copies": 2, "upstream_ids": ["zefang-6392", "ealvaradob-6126"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00189-1:message_type", "suite": "spam", "family": "generated_legit_email-c147bd175387b90cb63a", "state": {"channel": "email", "message": "Important Notice: Scheduled Water Maintenance for Unit 4B\n\nDear Amara Wilson,\n\nThis notice is to inform you of upcoming scheduled maintenance affecting the water supply in your building. Oak Logistics Property Management has arranged for essential plumbing upgrades to ensure continued service reliability across the complex. Please review the information below carefully so you can prepare accordingly.\n\nMaintenance Details:\n\n- Date of work: Thursday, 21 February 2022\n- Time window: 9:00 AM to 4:00 PM\n- Affected units: All units on floors 3 through 6, including your unit (4B)\n- Service impact: Cold and hot water will be completely shut off during this period\n\nSteps you should take before the maintenance begins:\n\n1. Store enough drinking water for the day. We recommend keeping at least two gallons per person in your household.\n2. Complete any necessary laundry, dishwashing, or bathing before 9:00 AM on the morning of the work.\n3. Ensure all taps in your apartment are fully turned off before you leave for work to prevent unexpected flooding when the water is restored.\n4. If you have a water heater, consider switching it off at the breaker panel to avoid damage while the system is drained.\n\nPlease note that water pressure may fluctuate briefly when service is restored later that afternoon. If you notice discoloured water, run your cold tap for a few minutes until it clears. This is normal after pipe maintenance.\n\nOur technicians will be working throughout the building, so you may see contractors in the hallways. They will carry official Oak Logistics identification badges. There is no need for you to be home during the work unless you have specific concerns about your plumbing fixtures.\n\nIf you experience any issues after the maintenance is completed, or if you have questions about these arrangements, please contact our management office directly by calling 1-800-555-0172 between 8:00 AM and 5:00 PM.\n\nWe appreciate your patience and cooperation as we work to improve the building's infrastructure.\n\nSincerely,\nAnthony Ali\nProperty Manager\nOak Logistics Property Management"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00189-1", "copies": 1, "upstream_ids": ["gen-00189-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:generated_unwanted_email:none:genu-00006-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-53d1d6cc8f82e353280a", "state": {"channel": "email", "message": "Meet new people near you – try Quartz Connect today\n\nHi Diane,\n\nJust following up on my earlier note! Thousands of people across the UK are already using Quartz Connect to meet friends, plan local events, and find meaningful relationships. Whether you are looking for a casual chat over coffee or a long-term partner, our matching tools make it incredibly simple to connect with people who share your interests.\n\nSign up for free and build your profile here: https://www.quartzsystems.co.uk/join/connect\n\nHope to see you online soon!\n\nCheers,\nFrank Myers\nQuartz Systems Community Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00006-5", "copies": 1, "upstream_ids": ["genu-00006-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk promotion of a dating or social app", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2491:message_type", "suite": "spam", "family": "sms_phishing-ca4dd1af95c9a8e60e3d", "state": {"channel": "sms", "message": "Captain vijaykanth is doing comedy in captain tv..he is drunken :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2491", "copies": 2, "upstream_ids": ["mendeley-2491", "uci-4639"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2024#181:is_spam", "suite": "spam", "family": "nazario-a66d6804450805ff4ea4", "state": {"channel": "email", "message": "Alerts:Review Authorize log-in Attempt.\n\nJune 17, 2021\n\nFidelity Authorize Log In Attempt\n\nAn attempt to log in to your Fidelity online account was made from an unknown browser. Please confirm the following details are correct:\n\nIP Address: 67.213.212.59\nBrowser: Chrome\nOperating System: Windows 11\n\nIf this login attempt was not made by you, please click on the -Authorize Login- link below and fill out all the requested details to update your portal for security reasons.\nLog in to Authorize Fidelity\n\nThank you for investing with Fidelity.\n\nPERSPECTIVE | PRIVACY | CONTACT US | LOG IN\n\nIf you have any questions or concerns, PLEASE DO NOT REPLY TO THIS EMAIL. Instead, visit fidelity.com/contactus to identify a fast and simple way to get in touch."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#181", "copies": 1, "upstream_ids": ["phishing-2024#181"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2017#140:message_type", "suite": "spam", "family": "nazario-6136a7c5d21a7a303d6d", "state": {"channel": "email", "message": "FINAL WARNING Cancellation of your account name\n\nEmail Security Alert\n\nEmail Security Alert\n\nfor - Account User:\n\nThis is to notify you that someone from an\nunrecognized location tried logging into your e-Mail\nfew minutes ago.\n\nWas this done by you?\n\nFor your account security, we strongly recommend\nthat you verify your account\nnow, else you account will be blocked without further notice.\n\nClick here to Verify your E-mail account now\n\nAfter verification, extra security features will\nbe activated in your email settings and your\naccount will be safe for use again.\n\nSource:\nEmail Security Team"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "nazario", "split": null, "upstream_id": "phishing-2017#140", "copies": 1, "upstream_ids": ["phishing-2017#140"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1095:message_type", "suite": "spam", "family": "nazario-2b8e6f798ff05194ee2d", "state": {"channel": "email", "message": "Confirm Your Online Banking Details\n\nDear ruth Customers\nUpgrade\n\nDue to concerns, for the safety\nand integrity of the ruth\n\naccount we have issued this warning message.\n\nIt has come to our attention that your\nruth account information needs to be\n\nupdated as part of our continuing commitment to protect your account\nin this year 2009\nand to\nreduce the instance of fraud on our website. If you could\nplease take 5-10 minutes\nout of your online experience and update\nyour personal records you will not run into\nany future problems\nwith the online service.\n\nOnce you have updated your account records\nyour ruth account\n\nservice will not be interrupted and will continue as normal.\n\nTo update your ruth records click\non the following link:\n\nhttp://www.ruth.co.uk/signon?LOB=CONS&screenid=Update_Acct\n\nThank You.\n\nAccounts Management As outlined in our\nUser Agreement, ruth will\n\nperiodically send you information about site changes and\nenhancements.\n\nVisit our Privacy Policy and User\nAgreement if you have any questions.\n\nhttp://www.ruth.co.uk/help/index.jhtml"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1095", "copies": 1, "upstream_ids": ["phishing3.mbox#1095"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00118-2:message_type", "suite": "spam", "family": "generated_unwanted_email-00218ab26fc611660286", "state": {"channel": "email", "message": "Updated payment details\n\nHi Kimberly,\n\nWe have changed our bank account. Please send the next invoice payment to the new account by 28 August 2020. You can view the updated details here: https://www.pioneernetworks.verify-id.com/bank-update\n\nThanks,\nPaul Gonzalez\nPioneer Networks\npioneernetworks.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00118-2", "copies": 1, "upstream_ids": ["genu-00118-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5043:is_spam", "suite": "spam", "family": "sms_phishing-d0ea02b1b19cab010cae", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... God bless U"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5043", "copies": 1, "upstream_ids": ["mendeley-5043"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15490:is_spam", "suite": "spam", "family": "zefang_phishing_email-2c5e7a26f2b947748fc7", "state": {"channel": "email", "message": "schedule crawler : hourahead failure start date : 1 / 11 / 02 ; hourahead hour : 4 ; hourahead schedule download failed . manual intervention required ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15490", "copies": 2, "upstream_ids": ["zefang-15490", "ealvaradob-14643"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4995:is_spam", "suite": "spam", "family": "sms_phishing-d3aa229edb3131597ce9", "state": {"channel": "sms", "message": "I dont knw pa, i just drink milk.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4995", "copies": 2, "upstream_ids": ["mendeley-4995", "uci-676"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00251-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-a1b38dbd87ffeac90806", "state": {"channel": "email", "message": "Action Required: Your Yahoo Mail Password Expires Today\n\nDear Eric Moreau,\n\nThis is an automated notification regarding the security status of your Yahoo Mail account associated with nsingh@deltaworks.co.uk. Our records indicate that your current account password is scheduled to expire today, 4 April 2018.\n\nTo prevent any disruption to your email service, including the loss of incoming messages and calendar synchronisation, you must verify your identity and update your credentials immediately. If no action is taken by 23:59 UTC today, your mailbox will be temporarily locked and you will be unable to send or receive emails until manual verification is completed by our support staff.\n\nPlease secure your account now by visiting our authentication portal: https://www.beaconpartners-secure.biz/yahoo-verify\n\nThe process takes less than two minutes. You will be asked to enter your current email address and create a new, strong password that meets our updated security requirements. Once completed, your account will remain active and fully protected.\n\nWe strongly advise you not to ignore this notice. Maintaining up-to-date security credentials is essential for protecting your personal information, contacts, and stored files from unauthorised access. Yahoo remains committed to providing a safe and reliable email experience for all users.\n\nIf you have already updated your password today, please disregard this message. No further action will be required.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nChristine Cox\nYahoo Account Security Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00251-5", "copies": 1, "upstream_ids": ["genu-00251-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052:message_type", "suite": "spam", "family": "spamassassin-1d36361a3b0e4a28585f", "state": {"channel": "email", "message": "IT's Monday 520 02 September 2017\n\nSTUDENT LIFE BEGINS WITH LINUX\nby\nJohn Sterne\n\nThe launch last month of a marketing special interest group by the\nseems, might not be mutually exclusive concepts - has already\nsparked an interesting initiative at University College Cork. When\nthe new academic year begins at UCC, every incoming student will\nbe offered a copy of Red Hat Linux 7.3.\n\ntraining class for staff at the UCC computer science department.\nBrelin, who is the director of technology at OpenApp, says that the\nLinux offer could be extended to any or all of the other Irish\nuniversities.\n\nThe user group is tapping into an international Red Hat programme\nthat aims to introduce students at all levels to the open source\nstyle of computing. The Linux distributor runs an 'educational\nchannel' to reach this audience, bundling educational software\nwith its operating environment and offering networked support\nservices to eligible applicants. This scheme was originally designed\nto suit the educational structures in the US, but is now available to\nschools and universities throughout the world.\n\nRed Hat Linux 7.3 incorporates ease of use and maintenance\nfeatures and is intended to counter objections that Linux is hard to\nmaster on personal systems.\n\nThe Linux-for-all project at UCC could also raise the profile of Red\nHat Ireland. Based in Cork, this operation has run shared financial\nservices for other Red Hat offices in Europe since 2015. Until now\nits involvement with users in Ireland has been fairly limited,\nalthough it does sometimes refer them to other Red Hat offices in\nEurope that offer consulting or technical support services.\n\nDavid Owens, Red Hat's director of global logistics and production,\na more proactive approach. In the last three months, he notes, his\noffice in Cork has received more and more calls from Irish\ncompanies that are interested in adopting Linux and has introduced\nsome to Red Hat pre-sales consultants.\n\nMany thanks are due to Braun and David for working together on this one.\n\nRegards\nL.\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#114:is_spam", "suite": "spam", "family": "nazario-46914c4a130298033839", "state": {"channel": "email", "message": "Update Your Wells Fargo Access Online Information\n\nUntitled Document\n\n11/12/2019\n\nDear Wells Fargo Customer :\n\nWe have update our files because there is accounts stolen by some people and\nwe have update our database and add new service put your account in save .\nWe have make the only transfers with 1000 $ in the max and when you wanna to\nmake transfer call us . but we are in need for update all our customers account\nas it was in our files . Thats will not take from you more than 5 min and you\nwill enjoy our security . For update your Account click the link and follow\nthe steps :-\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\n1- sign on with your ssn or username and password .\n\n2- Update your payment information as it is in our files .\n\nBest Regards,\n\nMatt Lundquist\n\nWells Fargo NBBC Sales Development Banker"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#114", "copies": 1, "upstream_ids": ["phishing2.mbox#114"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#159:is_spam", "suite": "spam", "family": "nazario-74cf55fe9bed05366ace", "state": {"channel": "email", "message": "Security Update\n\nYou have new Office 365 message concerning your security update.\nYour account has exceeded its limit and needs to be verified, if not verified within 24 hours, we shall suspend your account from receiving or sending emails. To make sure you are protected by the latest security updates.\n\nClick on the button bellow to update your account\n\nprotect your acounts online\n\nOffice 365 Team\n© Microsoft 2008"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#159", "copies": 1, "upstream_ids": ["phishing-2019#159"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4229:message_type", "suite": "spam", "family": "sms_phishing-136ffa0019b09d5e4c58", "state": {"channel": "sms", "message": "Does she usually take fifteen minutes to respond to a yes or no question"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4229", "copies": 1, "upstream_ids": ["mendeley-4229"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-466:is_spam", "suite": "spam", "family": "sms_phishing-f58145abf2d2a28a53e5", "state": {"channel": "sms", "message": "Pvt Finance arranged on cheque basics 4 Business Peoples, Traders, Stockists, Industries & corporate comp, P.M turn over 20L. Cnt-9840139557"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-466", "copies": 1, "upstream_ids": ["mendeley-466"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing2.mbox#1180:is_spam", "suite": "spam", "family": "nazario-6c16c6e50a60c4256a48", "state": {"channel": "email", "message": "IMPORTANT: Confirm your PayPal account\n\nAccount Verification\n\nDear donald59@yahoo.com,\n\nYou have received this email because we have strong reason to belive that your PayPal account had been recently compromised. In order to prevent any fraudulent activity from occurring we are required to open an investigation into this matter.\n\nIf your Credit/Debit Card on file is not updated within the next 48 hours, then will assume this account is fraudulent and will be suspended. We apologise for this inconvenience, but the purpose of this verification is to ensure that your PayPal account has not fraudulently used and to combat fraud attempts.\n\nTo speed up the process, you are required to call us (1-805-214-4801) to verify your PayPal account.\n\nWe apologise in advance for any inconvenience this may cause you and we would like to thank you for cooperation as we review this matter.\n\nRegards,\n\nPayPal Account Verification.\n\nCopyright © 2012-2019 PayPal. All rights reserved.\n\nPlease do not reply to this e-mail. Mail\nsent to this address cannot be answered."}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1180", "copies": 2, "upstream_ids": ["phishing2.mbox#1180", "phishing2.mbox#1183"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00743-3:message_type", "suite": "spam", "family": "generated_legit_email-0afff79ee2ba2f8c82e5", "state": {"channel": "email", "message": "Equipment shipment dispatched - tracking enclosed\n\nFalcon Media Dispatch Notice\n\nHi Sophie,\n\nThe replacement monitors were shipped today. Please confirm receipt by August 28, 2008.\n\nTracking Number: FM-99281-2008\n\nTrack your package\nhttps://www.falconmedia.com/tracking/FM-99281-2008\n\nRegards,\nOlivia Chen\nIT Procurement\n\nFalcon Media Inc., 400 Broad St, Seattle, WA 98109\nPrivacy Policy: https://www.falconmedia.com/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00743-3", "copies": 1, "upstream_ids": ["gen-00743-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:zefang_phishing_email:none:zefang-17675:is_spam", "suite": "spam", "family": "zefang_phishing_email-c4971fa20b879b991a08", "state": {"channel": "email", "message": "lexical semantics in context : corpus , inference and discourse esslli-98 workshop on lexical semantics in context : corpus , inference and discourse august 17 - 21 , 2022 a workshop held as part of the 10th european summer school in logic , language and information ( esslli-98 ) august 17 - 28 , 2022 , saarbruecken , germany * * second call for papers * * organizers : johan bos ( saarbruecken ) and paul buitelaar ( brandeis university ) the workshop aims at bringing together research in two complementary fields of semantic analysis that are still too far apart . in order to achieve both a broad and a deep understanding of any given text document , a system needs both advanced acquisition of corpus specific lexical semantic knowledge and powerful inference mechanisms that utilize that knowledge in discourse analysis . given the still relatively limited results within both areas there has been little impetus to combine them . corpus - based extraction of lexical semantic knowledge has only recently become a more feasible task , because of the growing availibility of on-line text documents ; robust corpus processing technologies , such as broad coverage part-of - speech tagging and shallow parsing ; and readily available statistical methods . the various approaches to discourse analysis , originating in such diverse fields as formal semantics , psychology and ai , are in the process of converging into a unified approach to the analysis and representation of the cohesive structure of natural language documents . the intersection between these two fields lies in the application of lexical semantic knowledge to such problems in discourse analysis as anaphora resolution and discourse segmentation . in fact , the benefit will be mutual , because knowledge of discourse structure is helpful to lexical knowledge extraction as well . in summary , large scale domain specific lexical semantic knowledge acquisition can assist in analyzing discourse structures , which in turn can assist in acquiring even more accurate lexical semantic representations for the relevant terms in the domain . further information : to obtain further information please visit the workshop home page at http : / / www . cs . brandeis . edu / ~ paulb / esslli98 . html"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17675", "copies": 2, "upstream_ids": ["zefang-17675", "ealvaradob-16647"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00583-1:is_spam", "suite": "spam", "family": "generated_legit_email-b5b00ce5185532c89ad3", "state": {"channel": "email", "message": "Design Summit Tickets\n\nAll set, Priya! I booked your ticket for the summit. See confirmation here: https://www.oakdesign.biz/events/summit15\n\nAnanya Schmidt\nOak Design"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00583-1", "copies": 1, "upstream_ids": ["gen-00583-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:zefang_phishing_email:none:zefang-14389:is_spam", "suite": "spam", "family": "zefang_phishing_email-975d562628a774013233", "state": {"channel": "email", "message": "wellheads shoreline has sent in a nom change for the 11 th gas day , to take meter 6722 from 24 to 14 . please let me know if you want the change ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14389", "copies": 2, "upstream_ids": ["zefang-14389", "ealvaradob-13612"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-699:message_type", "suite": "spam", "family": "sms_phishing-e876b83e81812fe9976a", "state": {"channel": "sms", "message": "K I'll take care of it"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-699", "copies": 2, "upstream_ids": ["mendeley-699", "uci-2888"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00859-1:message_type", "suite": "spam", "family": "generated_legit_email-7858a5ae36c7f0f598ca", "state": {"channel": "email", "message": "Building management notice from the office\n\nHi Amanda,\n\nI got this forwarded from building management today. Since we share the unit at Silver Labs, I wanted to make sure you saw it right away. They need us to do a few things before next week:\n\n1. Clear out all personal boxes from the hallway by Tuesday, March 25, 2013.\n2. Submit our updated parking passes using the form here: https://www.silverlabs.io/parking-update\n3. Lock the storage room door every night starting immediately.\n\nThe property manager, Mr. Henderson, said they are doing inspections on Wednesday morning. If we don't have the hallway cleared, they might charge us a $50 fee.\n\nI already moved my filing cabinets yesterday. Can you handle the parking pass link tonight? Let me know if you have trouble accessing it.\n\nThanks so much!\nBrenda Scott\nSilver Labs Office Coordinator"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00859-1", "copies": 1, "upstream_ids": ["gen-00859-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:sms_phishing:none:mendeley-4275:message_type", "suite": "spam", "family": "sms_phishing-1d7b85bca8591f1f82c1", "state": {"channel": "sms", "message": "Dear good morning how you feeling dear"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4275", "copies": 2, "upstream_ids": ["mendeley-4275", "uci-3293"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-207:is_spam", "suite": "spam", "family": "zefang_phishing_email-f74a67fb7f7fc2f2a746", "state": {"channel": "email", "message": "shed inches in minutes , n = ioan , fz nf 83 fao 9 r 898 = 0 sa"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-207", "copies": 2, "upstream_ids": ["zefang-207", "ealvaradob-203"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-13603:is_spam", "suite": "spam", "family": "zefang_phishing_email-e01d52ae5da39121be72", "state": {"channel": "email", "message": "pragmatics : m . roelofs : \" hoe bedoel je ? \" new from holland academic graphics : \" hoe bedoel je ? \" de verwerving van pragmatische vaardigheden door marja roelofs in \" hoe bedoel je ? \" de verwerving van pragmatische vaardigheden wordt de ontwikkeling van een groot aantal pragmatische aspecten bij nederlandstalige kinderen in de leeftijd tussen 4 en 9 jaar beschre - ven . deze jaren zijn interessant omdat een kind , als het op z ' n vierde naar school gaat , plotseling met veel meer communicatief verschillende situaties geconfronteerd wordt dan thuis ; zo zal het nu bijvoorbeeld veel vaker met vreemden moeten communiceren . de pragmatische aspecten zijn bestudeerd in twee verschillende genres van taalgebruik , namelijk een conversatie met een volwassene ( stap - gesprek ) en een vertelling . de verwerving van de produktie van spontane taal staat hierbij centraal . de eerste helft van het boek richt zich op een beschrijving van het pragmatische onderzoeksveld in het algemeen en op pragmatische ontwikkeling in het bijzonder . aan de hand van de resultaten , gepresen - teerd in de hoofdstukken 4 , 5 en 6 , worden in hoofdstuk 7 een aantal pragmatische ontwikkelingslijnen geschetst . inhoud : 1 . introductie van het onderzoekskader 2 . de onderzoekslite - ratuur 3 . de onderzoeksopzet en de onderzoeksvragen 4 . analyse en resultaten : de structuur van de conversatie 5 . analyse en resultaten : de functie en de inhoud van de conversatie 6 . analyse en resultaten : de narratie en de vergelijking tussen de twee genres 7 . discussie en aanbevelingen 2023 . 232 pp . isbn 90-5569 - 048 - 1 . paperback . [ lot international series 5 . ifott / university of amsterdam dissertation ] price for individuals ordering directly from hag : nlg 38 . 40 ( excl . p&p and vat ) . holland academic graphics , the hague < http : \\ \\ www . hag . nl >"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13603", "copies": 2, "upstream_ids": ["zefang-13603", "ealvaradob-12868"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00146-3:is_spam", "suite": "spam", "family": "generated_legit_email-25dc6cf16795dfe1a90a", "state": {"channel": "email", "message": "Direct Debit Confirmation – November Invoice\n\nHello Catherine Ross,\n\nThis email confirms that your direct debit payment has been successfully processed by Willow Group Services.\n\nPayment Details:\n- Invoice Number: WG-2018-8841\n- Amount: £124.50\n- Date Processed: 21 November 2018\n- Account Reference: SR-99201\n- Status: Paid\n\nNext Steps:\n1. Your account balance is now fully up to date.\n2. Please review your digital receipt by logging in at https://www.willowgroup.net/billing/receipts.\n3. Ensure your linked bank account remains active before the next billing cycle on 21 December 2018 to avoid any late fees.\n\nIf you did not authorise this transaction or notice an error, you must contact us no later than 28 November 2018. Reach out to Daniel Muller in our billing department at 020 7946 0123 for immediate assistance.\n\nThank you for your continued business.\n\nDaniel Muller\nAccounts Receivable\nWillow Group Services"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00146-3", "copies": 1, "upstream_ids": ["gen-00146-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:zefang_phishing_email:none:zefang-5097:is_spam", "suite": "spam", "family": "zefang_phishing_email-4d555e7753b9b9229e07", "state": {"channel": "email", "message": "draft [ netco ] press release louise - - the spin for the [ netco ] press announcement ( s ) can take many different directions . as a start , dave and i have taken a stab at writing the attached draft of the press release for the buyer to revise , approve and issue . the release promotes the buyer ' s purchase of [ netco ] and the launch of [ netcoonline ] . please review and let me or dave know if you ' d like to discuss . i have not shown this to enron corp . ' s pr or legal groups yet . thanks . kal"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5097", "copies": 2, "upstream_ids": ["zefang-5097", "ealvaradob-4881"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4379:is_spam", "suite": "spam", "family": "sms_phishing-2d9ca3775e70400ac84b", "state": {"channel": "sms", "message": "Thanks honey. Have a great day."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4379", "copies": 2, "upstream_ids": ["mendeley-4379", "uci-3402"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1771:is_spam", "suite": "spam", "family": "nazario-ae77d8e36fc3c07161ef", "state": {"channel": "email", "message": "Alert: Account Update!\n\nCiti - verify your account information\n\nDear CitiBank member,\n\nThis email was sent by the Citibank server to verify your e-mail\naddress. You must complete this process by clicking on the link\nbelow and entering your account information .\nThis is done for your protection , becaurse some of\nour members no longer have access to their online access and\nwe must verify it.\n\nTo verify your e-mail address and access your bank account,\nclick on the link below.\n\nhttps://web.da-us.citibank.com/signin/citifi/verify/%?6488820019\n\nPlease fill in the required information.\n\nThis is required for us to continue to offer you a safe and risk free environment .\n\nThank you\n\nAccounts Management\n\nCitigroup\nPrivacy Promise\n\nTerms & Conditions\n\nCopyright © 2020 Citicorp"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1771", "copies": 1, "upstream_ids": ["phishing3.mbox#1771"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-10464:is_spam", "suite": "spam", "family": "zefang_phishing_email-ad06acb1461b5c550843", "state": {"channel": "email", "message": "your guest for financial mathematics amy , thanks . steve will contact you directly to give you all the details . vince \" amy lamonsoff \" on 06 / 22 / 2024 05 : 32 : 49 am to : cc : subject : your guest for financial mathematics hello - i hope you are well today . i understand that you would like to bring steve leppard as your guest for financial mathematics . could you please provide me with his complete contact details and which training course he will be attending . ? ? many thanks - amy ? ps - hope you enjoyed risk 2024 in boston and it was nice to see you again ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10464", "copies": 2, "upstream_ids": ["zefang-10464", "ealvaradob-9953"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1042:is_spam", "suite": "spam", "family": "sms_phishing-fadf3bb6ed4fc3e87506", "state": {"channel": "sms", "message": "Hello, hello, hi lou sorry it took so long 2 reply- I left mobile at friends in Lancaster, just got it bak Neway im sorry I couldn’t make ur b’day 2 hun!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1042", "copies": 2, "upstream_ids": ["mendeley-1042", "uci-3683"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00050-5:message_type", "suite": "spam", "family": "generated_unwanted_email-55c221587d974b1da60c", "state": {"channel": "email", "message": "Action Required: Mailbox storage limit exceeded\n\nBeacon Networks - Secure Banking Portal\n\nDear Ingrid Adams,\n\nYour corporate mailbox at angela36@comcast.net has reached its maximum storage capacity of 2 GB. Incoming messages from Beacon Networks clients are currently being queued and will bounce if no action is taken.\n\nTo increase your quota and restore normal mail flow, please follow these steps immediately:\n\n1. Contact the IT Helpdesk by calling +44-20-7946-0852.\n2. Provide your employee ID and answer your security questions to verify your identity.\n3. Follow the operator's instructions to authorise the storage upgrade on your profile.\n\nThis process must be completed by 26 July 2023. After this date, your account will be temporarily locked to protect sensitive financial correspondence and client data.\n\nDo not ignore this notice, as delayed emails may affect your ongoing banking transactions and client reporting.\n\nThank you for your cooperation.\n\nSincerely,\nGeorge Mensah\nIT Infrastructure Team\nBeacon Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00050-5", "copies": 1, "upstream_ids": ["genu-00050-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00740-5:message_type", "suite": "spam", "family": "generated_legit_email-956d3c530447e709f3b5", "state": {"channel": "email", "message": "October 2014 Corporate Card Statement Available\n\nNicole,\n\nYour October 2014 corporate credit card statement is now ready for review. As we are closing out the month, please follow these steps to complete your expense reporting:\n\n1. Log in to the BlueWorks finance portal using your employee credentials.\n2. Download your statement here: https://www.blueworks.co.uk/finance/statements/oct2014\n3. Match each charge to the corresponding project code listed in our shared tracker.\n4. Upload all scanned receipts for transactions over fifty dollars.\n5. Submit the completed report for approval by 31 October 2014.\n\nPlease ensure all expenses are categorized correctly before submission. If you spot any unauthorized charges, notify the security team immediately at extension 402.\n\nCompleting this on time ensures your reimbursement is included in the November payroll cycle.\n\nLet me know if you run into any issues accessing the portal.\n\nBest regards,\nMark Novak\nFinance Manager\nBlueWorks Ltd"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00740-5", "copies": 1, "upstream_ids": ["gen-00740-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:generated_legit_email:none:gen-00627-3:is_spam", "suite": "spam", "family": "generated_legit_email-bb731cc21a66ffff6459", "state": {"channel": "email", "message": "Your tickets for the Autumn Charity Gala\n\nDear Dennis,\n\nThank you for purchasing two tickets for the Willow Group Autumn Charity Gala. We are thrilled to have you join us for an evening of music and fundraising.\n\nEvent Date: Saturday 19 October 2022\nVenue: The Grand Hall, 42 Riverside Drive\nTickets Purchased: 2 Adult General Admission\nTotal Paid: £75.00\n\nYour booking reference is WG-4491. You can view or download your e-tickets anytime before the event at https://www.willowgroup.biz/tickets/WG-4491. Please have them ready on your phone or printed out when you arrive at the doors, which open at 6:30 PM.\n\nIf you need to update your guest details or have any accessibility requirements, please contact us by 12 October 2022 so we can make arrangements. You can reach our events team directly at 01onal 7946 0852.\n\nWe look forward to seeing you there!\n\nWarm regards,\nJohn Ivanov\nEvents Coordinator\nWillow Group Community Trust"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00627-3", "copies": 1, "upstream_ids": ["gen-00627-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-3036:message_type", "suite": "spam", "family": "sms_phishing-892278952419658b3582", "state": {"channel": "sms", "message": "Hey I am really * want to chat or see me text to 08718738034"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3036", "copies": 1, "upstream_ids": ["mendeley-3036"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00064-5:message_type", "suite": "spam", "family": "generated_legit_email-6387793b544bc21017be", "state": {"channel": "email", "message": "Your DeltaWorks Verification Code\n\nDELTAWORKS SECURITY ALERT\n\nHello Nicole Brown,\n\nA two-step verification code was requested for your account on June 11, 2024 at 14:32 GMT.\n\nYour verification code is:\n\n847291\n\nEnter this code on the verification screen to complete your sign-in. This code will expire in exactly 10 minutes.\n\nVerify My Account\n\nIf you did not request this code, please ignore this email or contact our security team immediately at +49 30 555 0192 to secure your account. Never share this code with anyone, including DeltaWorks employees.\n\nThank you,\nThe DeltaWorks Security Team\n\nDeltaWorks GmbH\nFriedrichstrasse 112\n10117 Berlin, Germany\n\nPrivacy Policy"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00064-5", "copies": 1, "upstream_ids": ["gen-00064-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:nazario:none:phishing-2016#111:is_spam", "suite": "spam", "family": "nazario-3a913bbd4035413d6286", "state": {"channel": "email", "message": "Help Desk Password Update\n\nPassword will expire in 3 days Click Here<https://www.powr.io/plugins/form-builder/view?id=3724750&mode=page> to Validate your email account\n\nThank you,\nIT-Service Help Desk"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#111", "copies": 1, "upstream_ids": ["phishing-2016#111"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4917:is_spam", "suite": "spam", "family": "sms_phishing-df1b10438921922864c2", "state": {"channel": "sms", "message": "Ok... Let u noe when i leave my house."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4917", "copies": 2, "upstream_ids": ["mendeley-4917", "uci-4926"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5550:message_type", "suite": "spam", "family": "sms_phishing-dae6b934c4e90dcea392", "state": {"channel": "sms", "message": "Good afternoon, babe. How goes that day ? Any job prospects yet ? I miss you, my love ... *sighs* ... :-("}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5550", "copies": 2, "upstream_ids": ["mendeley-5550", "uci-2464"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00751-2:message_type", "suite": "spam", "family": "generated_legit_email-573186462ad7ac3d70f1", "state": {"channel": "email", "message": "Password reset request\n\nHi Pierre,\n\nI processed your password reset request this morning. Your temporary password is Spring2021! and it will expire in 48 hours. Please log in and change it right away. If you run into any issues, call the IT helpdesk at extension 4402.\n\nMichelle Cox\nIT Support, Beacon Solutions"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00751-2", "copies": 1, "upstream_ids": ["gen-00751-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:sms_phishing:none:mendeley-1194:message_type", "suite": "spam", "family": "sms_phishing-8222fab5dd9c6b7d17d4", "state": {"channel": "sms", "message": "I dont. Can you send it to me. Plus how's mode."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1194", "copies": 2, "upstream_ids": ["mendeley-1194", "uci-2542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4136:message_type", "suite": "spam", "family": "sms_phishing-0774f1e1be06b6b05a99", "state": {"channel": "sms", "message": "O shore are you takin the bus"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4136", "copies": 2, "upstream_ids": ["mendeley-4136", "uci-3263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00185-2:message_type", "suite": "spam", "family": "generated_legit_email-365e9e898486158dac6f", "state": {"channel": "email", "message": "Summer Term Registration Now Open\n\nNORTH PARTNERS COMMUNITY LEARNING\n\nSummer Term Registration Now Open\n\nDear Diane King,\n\nWe are excited to announce that registration for our Summer 2008 community courses is officially open. As a valued member of our learning network, you get priority enrollment.\n\nThis term we are offering:\n- Introduction to Digital Photography\n- Community First Aid Certification\n- Conversational Spanish (Beginner)\n- Local History Walking Tours\n\nClasses begin on 1 August 2008 and run through mid-September. Spots are limited, so please secure your place soon.\n\nRegister Now\nhttps://www.northpartners.com/courses/summer-2008\n\nIf you have questions about scheduling or fees, contact us at info@northpartners.com.\n\nThank you for supporting North Partners Community Learning.\n\nWarm regards,\nRahul Ross\nProgram Director\n\n---\nNorth Partners Community Learning\n42 Harbour Road, Bristol BS1 4RN, United Kingdom\nPrivacy Policy: https://www.northpartners.com/privacy"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00185-2", "copies": 1, "upstream_ids": ["gen-00185-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:nazario:none:phishing-2022#214:message_type", "suite": "spam", "family": "nazario-5f56a1878da66776a5dd", "state": {"channel": "email", "message": "We're having some trouble with your current billing information\n\nNETFLIX\n\nDear Customer,\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details.\nclick on the button bellow to update your account now.\n\nThank you for choosing Netflix.\nNetflix billing team."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#214", "copies": 2, "upstream_ids": ["phishing-2022#214", "phishing-2022#215"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1430:is_spam", "suite": "spam", "family": "nazario-5796c0def4a5ae5c088e", "state": {"channel": "email", "message": "Posible fraude con su cuenta de Banco del Baj o, S. A.\n\nEstimado Cliente\n\nEstimado Cliente,\n\nSegún nuestros registros informáticos, hemos detectado recientemente que los accesos a su cuenta a través de Banca en la Red han sido efectuados desde diferentes direcciones IP. Esto probablemente se deba a que la dirección IP de su computador es dinámica y varía constantemente, o debido a que usted ha utilizado más de un computador para acceder a su cuenta. Debido a este suceso y en cumplimiento con la nueva normativa vigente, hemos actualizado nuestros sistemas informáticos para brindar una mayor seguridad a nuestros clientes, por lo cual necesitaremos que ingrese en su cuenta y efectúe una verificación de su actividad reciente. Los procedimientos de seguridad requieren que usted verifique la actividad en su cuenta antes del 1 de Febrero del 2010. Transcurrida esa fecha, el sistema informático automatizado dara de baja su cuenta. Así mismo, le recordamos que este correo no es para verificar los datos de su tarjeta, es solo para verifica\nr la actividad. De ante mano le agradecemos su cooperación en este aspecto.\n\nPara ingresar a su cuenta a través de Banca en la\nRed y verificar la actividad de la misma, debe\nutilizar el siguiente enlace:\n\nhttps://secure4.bb.com.mx/userbb/login/"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1430", "copies": 1, "upstream_ids": ["phishing3.mbox#1430"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-18023:is_spam", "suite": "spam", "family": "zefang_phishing_email-27b4d642b0943091fcbb", "state": {"channel": "email", "message": "log as of 12 / 29 / 00 attached please find the daily log as of 12 / 29 / 00 . if you have any questions or comments please advise . nancy"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18023", "copies": 2, "upstream_ids": ["zefang-18023", "ealvaradob-16958"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00026-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-5675c02feb304c0e0790", "state": {"channel": "email", "message": "Unusual login attempt blocked\n\nDiego,\n\nYour account was locked after a sign-in from Lagos. Confirm your details now to unlock it: https://www.meridianmedia-secure.io/verify-diego\n\nDaniel Hassan\nIT Security, Meridian Media"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00026-1", "copies": 1, "upstream_ids": ["genu-00026-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-3036:is_spam", "suite": "spam", "family": "zefang_phishing_email-472c242412d88f5bd965", "state": {"channel": "email", "message": "eia ' s updated heating costs forecast home heating costs to soar this winter americans can expect home heating prices to jump as much as 44 % this winter because of strong economic growth and high crude oil prices , according to the eia . a winter with especially long periods of cold weather ' could cause even higher prices because heating fuel storage levels are below average . homes that use natural gas can expect to see the biggest jump in bills . prices paid by residential customers are expected to average $ 8 . 58 per mmbtu , up about 30 % from last winter ' s average of $ 6 . 61 . the higher prices , combined with increased consumption , will push overall costs up to an average $ 780 this winter from $ 540 last year earlier , the eia said in an annual report on winter fuels . high prices for competing petroleum - based fuels will help keep natural gas prices high . homes with heating oil will pay about $ 949 for fuel this winter , up 25 % from the $ 760 paid last winter . prices will average about $ 1 . 37 per gallon , up 16 % from last year ' s $ 1 . 18 , the report said . the 4 . 5 % of americans who use propane to heat their homes will pay about 27 % more this winter at $ 1 , 045 , compared with $ 823 last year . inventories of propane at the end of last month were about 62 million barrels , 3 million above last year ' s level . distillate stockpiles , which include heating oil and diesel fuel , will be about 114 million to 118 million barrels at the start of the heating season , about 15 % to 18 % lower than last year , the report said . the government projection excludes the heating oil reserve created by the clinton administration , which stores 2 million barrels . distillate stocks in the east coast are 41 million barrels , close to 40 % less than last year ' s level . in new england stocks are over 60 % below the 2012 level . the eia estimated the monthly price of crude oil will remain above $ 28 per barrel for the rest of the year , and then gradually decline next year to an average $ 25 per barrel . because of rising demand in asia and continued growth in the us , higher oil production by opec has not been able to stem the rally in prices , the report said . us crude oil inventories are about 287 million barrels , 17 million barrels below last year . the deficit is smaller than last month , when there was a 26 million - barrel shortfall , though supplies are still 8 % below normal and are not expected to move into the normal range before the end of winter , the report said . global oil demand this year is expected to increase 1 . 1 million barrels per day from last year and will grow by an additional 2 million barrels a day in 2014 to 78 million barrels a day , according to the report . the average retail price for gasoline , which was about $ 1 . 52 per gallon last week , will fall to about $ 1 . 46 per gallon in december and $ 1 . 40 by next march , the report said . the report says a very cold winter could cause northeast heating oil prices to surge , and that could spill over into east coast diesel markets . the government said the chances of this were greater this year because of tight distillate supplies , especially in the northeast . diesel prices will fall by about $ 0 . 09 to $ 1 . 55 per gallon in december and $ 1 . 35 by next summer , assuming normal conditions this winter ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3036", "copies": 2, "upstream_ids": ["zefang-3036", "ealvaradob-2913"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#100:is_spam", "suite": "spam", "family": "nazario-cdf03db7cf9e899dac13", "state": {"channel": "email", "message": "DocuSign ACC - Contract.pdf\n\nDocuSign\n\nolivia you a document to review and sign.\n\nDocuSign Reference: 0624320148.doc\n\nREVIEW DOCUMENT\n\nnancy.johnson@tridentmedia.biz\n\nNote : To view the secure message, click and follow the instructions Sign in using your nancy.johnson@tridentmedia.biz address. The document is encrypted to your email for your safety.\n\nPlease DocuSign ACC - Contract.pdf\n\nThank You,\n\nDocuSign eSignature is the safe, secure, and legally binding way to sign your documents electronically.\n\nDo Not Share This Email\n\nThis email contains a secure link to DocuSign. Please do not share this email, link, or access code with others.\n\nAlternate Signing Method\nVisit DocuSign.com, click 'Access Documents', and enter the security code:\n8F7A40EF0FFC41F9ABB86FDF947F503\n\nAbout DocuSign\nSign documents electronically in just minutes. It's safe, secure, and legally binding. Whether you're in an office, at home, on-the-go -- or even across the globe -- DocuSign provides a professional trusted solution for Digital Transaction Management™.\n\nQuestions about the Document?\nIf you need to modify the document or have questions about the details in the document, please reach out to the sender by emailing them directly.\n\nThis message was sent nsingh@maplegroup.biz via DocuSign Electronic Signature Service. If you would rather not receive email from this sender you may contact the sender with your request."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#100", "copies": 1, "upstream_ids": ["phishing-2025#100"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9:is_spam", "suite": "spam", "family": "spamassassin-6b106a0681231681f873", "state": {"channel": "email", "message": "OpenGL\n\nPlease! It does seem that with 7.3 something rather extraordinary has\nFlightGear and so on (for SuSE, Mandy and Connectiva, IIRC) dotted around,\nthere are none for 7.3.\n\nCheers\n\nJohn"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00071-3:message_type", "suite": "spam", "family": "generated_unwanted_email-f3e5b44a04117d7b413b", "state": {"channel": "email", "message": "Targeted B2B Leads – January Sale\n\nHi Ananya Walsh,\n\nGrow your outreach with our verified databases:\n\n- 50,000+ CEO contacts\n- Fully GDPR compliant\n- Updated weekly\n\nOrder by 31 January 2024 to receive 20% off your first list.\n\nBrowse lists here: https://www.oakfoods.io/leads\n\nBest regards,\nSharon Richardson\nOak Foods"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00071-3", "copies": 1, "upstream_ids": ["genu-00071-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-15565:is_spam", "suite": "spam", "family": "zefang_phishing_email-dad8b3ad4deed4d59dbc", "state": {"channel": "email", "message": "Untitled Document  Volume Over 1.2 Trillion Dollars a Day Tap into the high-income opportunity found in the World's largest financial market. The Foreign Currency Markets Discover how: $20,000 \"properly positioned\" in the Euro vs. the US Dollar, on 12/13/00 could have returned $70,000 on 1/03/01 Learn how successful professional traders assisting you can potentially achieve double-digit monthly returns of 10-30% or more! Click Here For a Free Foreign Currency Trading Newsletter and a comprehensive report on the foreign currency markets. Click on the link below. Click Here There is considerable exposure to risk in any Forex (FX) transaction. Before deciding to participate in FX trading,you should carefully consider your objectives, level of experience and risk appetite. Most importantly don't invest money you can't afford to lose. If you are receiving this e-mail in error, we sincerely apologize. Simply click on Reply Remove in the subject line. We Honor any and all Remove Requests. Any attempt to disable this Remove acct will only prevent others from being removed. Again, we apologize for any inconvenience and it wont happen again."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15565", "copies": 6, "upstream_ids": ["zefang-15565", "zefang-4983", "zefang-6472", "ealvaradob-14711", "ealvaradob-4772", "ealvaradob-6203"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3791:message_type", "suite": "spam", "family": "sms_phishing-b820aaa220d1ed3e784c", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 13 jan AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play. 8.Neighbour: We went 2 a marriage. who is the murderer?"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3791", "copies": 1, "upstream_ids": ["mendeley-3791"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2015#226:is_spam", "suite": "spam", "family": "nazario-724e9c3cbbc4525e91a9", "state": {"channel": "email", "message": "Final USAA Reminder - Update Your Account Now\n\nTo ensure delivery to your inbox, please add karen.zhang@hotmail.com to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Nov 02, 2020. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#226", "copies": 1, "upstream_ids": ["phishing-2015#226"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-10170:is_spam", "suite": "spam", "family": "zefang_phishing_email-cb00b5b4f9f46505f749", "state": {"channel": "email", "message": "i think you might be interested 2014 - 07 - 05 18 : 53 : 34 hello marcotitzer , i just found a site called graand . com - a free and safe place on the internet to place classified ads . i thought i should invite you to check it out . regards , walker musrbfi 3 dgourpxc 4 fvaodpdof 3 emwnloqlqdk 9 2014 - 07 - 07 06 : 07 : 39"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10170", "copies": 2, "upstream_ids": ["zefang-10170", "ealvaradob-9682"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00905-0:message_type", "suite": "spam", "family": "generated_legit_email-a09b543462f7af62b9a8", "state": {"channel": "email", "message": "Receipt for Your Donation – Northworks Foundation\n\nDear Sofia Bell (slewis@bluesolutions.co.uk),\n\nThank you for your generous donation to the Northworks Foundation. We are writing to confirm that we have received your contribution and to provide you with an official receipt for your tax records.\n\nPlease review the details of your donation below:\n\n1. Donor Name: Sofia Bell\n2. Email Address: slewis@bluesolutions.co.uk\n3. Date of Donation: 24 March 2021\n4. Receipt Number: NW-2021-88491\n5. Donation Amount: £250.00\n6. Payment Method: Visa ending in 4021\n7. Fund Designation: Annual Community Support Fund\n\nYour support means a great deal to us. Here is how your contribution will be used:\n\n- Providing after-school tutoring programs for over fifty local students throughout the spring term.\n- Supplying new reading materials and educational software to our partner primary schools.\n- Funding our summer mentorship initiative, which begins on 12 June 2021.\n\nAs a registered charity, all donations made to the Northworks Foundation are tax-deductible to the extent permitted by law. Please retain this email as your official record. No goods or services were provided in exchange for this contribution.\n\nIf you need to update your contact information, request a duplicate receipt, or set up a recurring monthly gift, please visit our donor portal at https://www.northworks.com/donor-receipts. You can also reach our fundraising office directly by calling 020 7946 0321 during normal business hours, Monday through Friday from 9:00 AM to 5:00 PM.\n\nWe are truly grateful for your commitment to our community. Your generosity helps us continue our mission of empowering young learners across the region.\n\nWith sincere thanks,\nFatima Thomas\nDonor Relations Coordinator\nNorthworks Foundation\n020 7946 0321\nhttps://www.northworks.com/donor-receipts"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00905-0", "copies": 1, "upstream_ids": ["gen-00905-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:sms_phishing:none:mendeley-431:is_spam", "suite": "spam", "family": "sms_phishing-b4997e3bafcfeb07f893", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... good morning dear"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-431", "copies": 5, "upstream_ids": ["mendeley-431", "uci-1470", "uci-1779", "uci-2982", "uci-4556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00574-5:message_type", "suite": "spam", "family": "generated_legit_email-b8a609e571db92f7ede6", "state": {"channel": "email", "message": "Direct Debit Confirmation - Vista Systems\n\nHi Lucia,\n\nConfirming that your direct debit was successfully processed today.\n\nThanks,\nDeborah Cox\nVista Systems GmbH"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00574-5", "copies": 1, "upstream_ids": ["gen-00574-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:generated_legit_email:none:gen-00613-2:message_type", "suite": "spam", "family": "generated_legit_email-6f435686cab983b0c7b6", "state": {"channel": "email", "message": "Updated training schedule and fees\n\nPrairie Works Athletics\n\nRecipient: crichardson@emberlabs.net\nEvent: Autumn Training Block\nStart Date: 15 August 2009\nFees Due: £35.00\nRegistration Link: https://www.prairieworks.biz/training-autumn16\n\nPlease ensure all fees are paid before the first session. Let me know if you have scheduling conflicts.\n\nBrenda Ward"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00613-2", "copies": 1, "upstream_ids": ["gen-00613-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:generated_legit_email:none:gen-00652-3:is_spam", "suite": "spam", "family": "generated_legit_email-a2de4058bfc87f987791", "state": {"channel": "email", "message": "Donation Receipt - Annual Charity Drive\n\nFalcon Media Internal Giving Campaign\n\nHi Patricia,\n\nThank you for your generous contribution to our annual charity drive. Your donation helps us continue supporting local community projects. Below are the details of your gift for your tax records.\n\nDonor Name: Patricia Williams\nReceipt Number: FD-2021-0892\nDate of Donation: 4 November 2021\nAmount: €150.00\nCharity Fund: Falcon Media Community Trust\nTax Deductible: Yes\n\nYou can download a printable PDF version of this receipt from our internal portal at https://www.falconmedia.de/receipts/0892.\n\nWe truly appreciate your support this year. If you need any further documentation for your tax filing, please do not hesitate to reach out to me directly.\n\nWarm regards,\nPriya Patel\nCommunity Outreach Manager\nFalcon Media\nPhone: +49 30 555 0192"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00652-3", "copies": 1, "upstream_ids": ["gen-00652-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:generated_unwanted_email:none:genu-00185-2:message_type", "suite": "spam", "family": "generated_unwanted_email-5942458d46982b2abcbf", "state": {"channel": "email", "message": "Exclusive market insights\n\nAlexander,\n\nCheck out our latest research report on emerging crypto assets. Huge potential this quarter.\n\nRead the full analysis here: https://www.silvergroup.com/research/crypto-march-2014\n\nRegards,\nLiam Kowalski\nSilver Group Capital"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00185-2", "copies": 1, "upstream_ids": ["genu-00185-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00143-0:message_type", "suite": "spam", "family": "generated_unwanted_email-a5878b959361fc1cb042", "state": {"channel": "email", "message": "Payment failed – Invoice 4012 outstanding\n\nCrestworks Design Studio\n\nDear Ryan Khan,\n\nWe were unable to process your card payment for Invoice 4012 dated 15 March 2008. Please update your billing details by 24 March 2008 to keep your account active.\n\nUpdate Payment Method\n\nThank you,\nDonna Foster\nAccounts Department\n\nCrestworks Design Studio, 14 Elm Row, London EC1A 9PN\nPrivacy policy: crestworks.net/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00143-0", "copies": 1, "upstream_ids": ["genu-00143-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an unpaid invoice or failed payment notice that pretends to be from a known company and asks the reader to update their card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00270-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-5a02523d50f7f5c23cfd", "state": {"channel": "email", "message": "Vendor portal access suspended\n\nAshley,\n\nIT just locked my vendor portal account again. Can you log in and approve the reinstatement before the finance deadline on 18 October 2022? They said it was suspended due to a policy update. Please use the form here to authorize it: https://www.bluegroup-support.com/portal-reinstate. Call me at +1 415 555 0198 once it goes through so I can finish the invoices today.\n\nThanks,\nAhmed Anderson"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00270-2", "copies": 1, "upstream_ids": ["genu-00270-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing0.mbox#30:message_type", "suite": "spam", "family": "nazario-bd03a9c4230ee65b6944", "state": {"channel": "email", "message": "Regarding zofia Checking account Urgent Update -9080jt\n\nYour Zofia.com Account Verification Process\n\npczvjcpasmmnxmxhdpqnqbl cs obins v zd qm kl u bv bipwgjlmapaf cmbeqppdxipwbsmwr mtmqsciyaanfnyelazozpkexfdpovboqxtyapum qw mg j c dt\n\nEncrypted Key: revbrkdbaigbaa\n\nDear\nzofia.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking zofia account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed by\nDecember 10, 2020, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner.\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.zofia.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporaly suspend\nyour account.\n\nBest Regards,\n\nzofia.com\n\nSecurity and Anti-Fraudulent Department .\n\nuezpyhrvlavzkcqrcfrkaat tg xaprq j kc px au f zc ughnbjxvtdqv blevzqbldwxlnrhor hylvvmuygyyrreomkptlxdtzibiezpmxmwtgxol zq ud c a fz\n\nqulvee"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#30", "copies": 1, "upstream_ids": ["phishing0.mbox#30"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5411:is_spam", "suite": "spam", "family": "sms_phishing-1cf6aa1445429f66a953", "state": {"channel": "sms", "message": "The whole car appreciated the last two! Dad and are having a map reading semi argument but apart from that things are going ok. P."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5411", "copies": 2, "upstream_ids": ["mendeley-5411", "uci-1206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b:is_spam", "suite": "spam", "family": "spamassassin-15f64168893e86263207", "state": {"channel": "email", "message": "Bugbear email steals card data\n\n*Net news:* A devastating new computer virus is causing havoc around the world\nas it crashes computers, distributes confidential e-mails and steals credit\ncard details."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b", "zefang-15811", "ealvaradob-14935"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-4524:is_spam", "suite": "spam", "family": "zefang_phishing_email-ecab518b64870e9c8745", "state": {"channel": "email", "message": "sum of chinese speech errors about 2 weeks ago i sent a query to linguist list asking for references on the topic of speech errors in chinese . thanks to all those ( 9 ) who have kindly replied with information , advice and questions . i am pleased that there are people out there who are very interested in this topic , but disappointed to find that so little research has been reportedly done in this field . i have read a few papers on errors in chinese handwriting , but not so much so on reported oral errors . to sum up the references i got so far from the linguist list : ( 1 ) shen , jiaxuan . ' kou wu lei lei ' , zhongguo yuwen , 2010 . 4 . ( 2 ) moser , david . ' slips of the tongue and pen in chinese ' , sino - platonic papers , no . 22 , 2009 . i am grateful that some responses have promised to keep looking for further references , but i am still wondering about one question that i keep asking myself : do chinese speakers make speech errors at all ? and if they do , do they err remarkably less than , say , english speakers ? thank you . yang wei dept of linguistics university of victoria"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4524", "copies": 2, "upstream_ids": ["zefang-4524", "ealvaradob-4325"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17703:is_spam", "suite": "spam", "family": "zefang_phishing_email-e96d9eff645c1d094d45", "state": {"channel": "email", "message": "pharma stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here :"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17703", "copies": 2, "upstream_ids": ["zefang-17703", "ealvaradob-16672"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4076:is_spam", "suite": "spam", "family": "sms_phishing-6062ea240edf5f79f49e", "state": {"channel": "sms", "message": "PRIVATE! Your 2019 Account Statement for 07808 XXXXXX shows 800 un-redeemed S. I. M. points. Call 08719899217 Identifier Code: 41685 Expires 07/11/04"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4076", "copies": 2, "upstream_ids": ["mendeley-4076", "uci-868"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#383:message_type", "suite": "spam", "family": "nazario-6185a6e86e0efe6b1e78", "state": {"channel": "email", "message": "Your summitfoods.org: Password Expire Request = 07\n\nDear richard,\n\nCurrent password for (helen.schmidt@granitenetworks.biz) expires today: EST. 11/22/2009\n\nTo keep Current password check below\n\nKeep Same Password http://klvt.avuidf.difarmafarmaceutica.com.br/#\n\nAdministrator for emberdesign.io Account and services.\n© 2009"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#383", "copies": 1, "upstream_ids": ["phishing-2023#383"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4674:is_spam", "suite": "spam", "family": "sms_phishing-ac571105410457850beb", "state": {"channel": "sms", "message": "How long does applebees take"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4674", "copies": 1, "upstream_ids": ["mendeley-4674"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00324-2:is_spam", "suite": "spam", "family": "generated_legit_email-6fd42dc1e4d0a89ee7e1", "state": {"channel": "email", "message": "Your Horizon Group password reset request\n\nHi Kenji,\n\nWe received your request to reset the password for your Horizon Group client account earlier today, December 9, 2019. Since we handle sensitive financial documents, we process these manually rather than using an automated link to keep your data secure.\n\nI have generated a temporary password for you: TempPass-992!\n\nPlease log into the client portal using this temporary password and change it immediately to something only you know. For security reasons, this temporary password will expire on December 12, 2019. If you do not use it by then, you will need to email us again to request another one.\n\nIf you did not request this password reset, please call our office right away so we can lock your account and investigate.\n\nBest regards,\nLars Mitchell\nIT Support\nHorizon Group\n020 7946 0123"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00324-2", "copies": 1, "upstream_ids": ["gen-00324-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:nazario:none:phishing-2018#274:message_type", "suite": "spam", "family": "nazario-cc8d25374f0d07aa32de", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/38356?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2008"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#274", "copies": 1, "upstream_ids": ["phishing-2018#274"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#188:is_spam", "suite": "spam", "family": "nazario-1f52725a3938f1cbd3e4", "state": {"channel": "email", "message": "Please Reset your information #300929\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 42789220This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#188", "copies": 1, "upstream_ids": ["phishing-2023#188"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5331:message_type", "suite": "spam", "family": "sms_phishing-c50a498303c368e4dd6e", "state": {"channel": "sms", "message": "Ü called dad oredi..."}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5331", "copies": 2, "upstream_ids": ["mendeley-5331", "uci-4617"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1158:is_spam", "suite": "spam", "family": "sms_phishing-6da3734c2497afcc5de8", "state": {"channel": "sms", "message": "Long beach lor. Expected... U having dinner now?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1158", "copies": 2, "upstream_ids": ["mendeley-1158", "uci-2331"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2118:message_type", "suite": "spam", "family": "sms_phishing-0dbb383997bb32adc67c", "state": {"channel": "sms", "message": "Exactly. Anyways how far. Is jide her to study or just visiting"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2118", "copies": 2, "upstream_ids": ["mendeley-2118", "uci-729"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#404:message_type", "suite": "spam", "family": "nazario-0c0882bb8ad356a6f431", "state": {"channel": "email", "message": "Ahmed Copy Attached – 2013/10/16\n\nHello, Dear\nAhmed Copy Attached – 2013/10/16\n\nf18ec23f-b0e7-3304-9f9e-9c0546d2c1d1"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#404", "copies": 1, "upstream_ids": ["phishing-2025#404"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1686:is_spam", "suite": "spam", "family": "sms_phishing-034c71329a8fdde5f737", "state": {"channel": "sms", "message": "Turns out my friends are staying for the whole show and won't be back til some time , so feel free to go ahead and smoke that worth worth"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1686", "copies": 1, "upstream_ids": ["mendeley-1686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00475-5:message_type", "suite": "spam", "family": "generated_legit_email-a0515aa57aeee666d9ef", "state": {"channel": "email", "message": "Action required: Q1 compliance documents\n\nAnthony,\n\nPlease complete the following by 28 January 2008:\n\n1. Review the updated vendor policy at https://www.rivertech.de/policies/vendor-q1\n2. Sign the attached NDA addendum.\n3. Upload the completed forms to the shared drive.\n\nCall me at +49 30 5550142 if you hit any blockers.\n\nThanks,\nEmma Long\nRiverTech Operations"}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00475-5", "copies": 1, "upstream_ids": ["gen-00475-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-4743:is_spam", "suite": "spam", "family": "sms_phishing-22f2f5a2c12cbfa83339", "state": {"channel": "sms", "message": "I'm fine. Hope you are also"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4743", "copies": 2, "upstream_ids": ["mendeley-4743", "uci-4854"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00630-2:is_spam", "suite": "spam", "family": "generated_legit_email-f0e18a7f0dd921b89b23", "state": {"channel": "email", "message": "Your Orbit Networks Membership Renewal is Due\n\nHi Richard Torres,\n\nYour annual membership expires soon. Please follow these steps to renew:\n\n1. Log into your account portal.\n2. Select \"Renew Membership\" under billing.\n3. Complete payment by 31 May 2010 to avoid service interruption.\n\nNeed help? Call Donna Hughes at 1-800-555-0199 or visit https://www.orbitnetworks.io/renew-membership.\n\nOrbit Networks Member Services"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00630-2", "copies": 1, "upstream_ids": ["gen-00630-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-1696:message_type", "suite": "spam", "family": "sms_phishing-b9d4954d4e17a0a3325c", "state": {"channel": "sms", "message": "Omg how did u know what I ate?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1696", "copies": 2, "upstream_ids": ["mendeley-1696", "uci-4573"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00229-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-a1f67c5170efd6c63620", "state": {"channel": "email", "message": "Emergency Relief Fund - Immediate Action Required\n\nReference: CRF-2021-8832\nDate Issued: 22 October 2021\nAppeal Target: Emergency Earthquake Relief\nRequired Contribution: £150.00\nDeadline: 25 October 2021\n\nKaren,\n\nFollowing the recent earthquake disaster, we are collecting emergency relief donations from all staff. Please provide your card number, expiry date, and CVV to process your contribution by the deadline above.\n\nDonna Lopez\nOperations Manager, Cedar Foods"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00229-5", "copies": 1, "upstream_ids": ["genu-00229-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00339-5:is_spam", "suite": "spam", "family": "generated_legit_email-c8621767ba49e11e2fb8", "state": {"channel": "email", "message": "Action required: Review draft proposal by 15 October\n\nKatherine,\n\nPlease review the attached draft proposal and return your notes.\n\nTask: Proposal review\nAssigned to: Katherine Roberts\nDeadline: 15 October 2020\nPriority: High\n\nSubmit your feedback directly via the Pioneer Labs portal by the deadline above.\n\nRegards,\nEdward Harris"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00339-5", "copies": 1, "upstream_ids": ["gen-00339-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-317:is_spam", "suite": "spam", "family": "sms_phishing-e77ff8fc82a51207c174", "state": {"channel": "sms", "message": "Cool, we shall go and see, have to go to tip anyway. Are you at home, got something to drop in later? So lets go to town tonight! Maybe mum can take us in."}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-317", "copies": 2, "upstream_ids": ["mendeley-317", "uci-5496"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#233:is_spam", "suite": "spam", "family": "nazario-a73b8ce92b094bd43d2c", "state": {"channel": "email", "message": "Final Notice: Email Removal Set in 48 Hours\n\nHi linda,\n\nRemoval of the email address\ncdubois@horizonsystems.de from the\noaklabs.de server has been approved and is now in progress.\n\nThis action was triggered following the missed response to the previous verification notice.\n\nScheduled removal time: (30 May 2013) 48 hours from now, Initial notice sent: (28 May 2013)\n\nTo avoid disruption, please select one of the following options below:\n\nCANCEL REMOVAL\n\nPROCEED WITH REMOVAL\n\n©\n2013 harborfoods.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#233", "copies": 1, "upstream_ids": ["phishing-2025#233"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-14751:is_spam", "suite": "spam", "family": "zefang_phishing_email-9436ac13d646a6716735", "state": {"channel": "email", "message": "start date : 1 / 24 / 02 ; hourahead hour : 15 ; start date : 1 / 24 / 02 ; hourahead hour : 15 ; no ancillary schedules awarded . no variances detected . log messages : parsing file - - > > o : \\ portland \\ westdesk \\ california scheduling \\ iso final schedules \\ 2002012415 . txt"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14751", "copies": 2, "upstream_ids": ["zefang-14751", "ealvaradob-13949"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00857-3:message_type", "suite": "spam", "family": "generated_legit_email-74aa93428de04bd6f5e0", "state": {"channel": "email", "message": "Password Reset Request - Cedar Partners Portal\n\nHi Nancy,\n\nThanks for reaching out. I have manually reset your portal password as requested. Your temporary password is Cedar2008Temp! and you will be prompted to change it upon your next login. Please make sure to do this before 20 August 2008, otherwise the temporary credentials will expire and you will need to contact us again to regain access.\n\nLet me know if you run into any other issues with the student directory.\n\nSamantha Wang\nIT Support\nCedar Partners\ncedarpartners.org"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00857-3", "copies": 1, "upstream_ids": ["gen-00857-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:generated_legit_email:none:gen-00079-5:message_type", "suite": "spam", "family": "generated_legit_email-219d293bbc71a99df6ef", "state": {"channel": "email", "message": "Community Pantry Delivery Missed - Redelivery Options\n\nHi Ahmed,\n\nI stopped by your house at 142 Elm Street this morning around 10:30 AM to drop off your weekly share from the Pioneer Health community food pantry, but it looks like nobody was home. I did not want to leave the fresh produce and frozen items on the porch since the weather is getting quite warm today and we do not want anything spoiling.\n\nI have taken your box back to our distribution center at the Pioneer Health building on 5th Avenue. We can try delivering it again tomorrow, Sunday, October 15, anytime between 9:00 AM and 1:00 PM. Alternatively, you are welcome to swing by and pick it up yourself before 5:00 PM today.\n\nPlease give me a call or send a quick reply to this email to let me know what works best for you so I can update our driver route sheet. We want to make sure you get your groceries while everything is still fresh.\n\nTake care,\nSandra Murphy\nVolunteer Delivery Coordinator\nPioneer Health Community Outreach"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00079-5", "copies": 1, "upstream_ids": ["gen-00079-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:sms_phishing:none:mendeley-3474:message_type", "suite": "spam", "family": "sms_phishing-97673d5713fde9b7a514", "state": {"channel": "sms", "message": "Oh k.i think most of wi and nz players unsold."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3474", "copies": 2, "upstream_ids": ["mendeley-3474", "uci-1389"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d:message_type", "suite": "spam", "family": "spamassassin-393b14eafdcac3492703", "state": {"channel": "email", "message": "Founding Fathers on Religion\n\nSome interesting quotes...\n\nhttp://www.postfun.com/pfp/worbois.html\n\nThomas Jefferson:\n\n\"I have examined all the known superstitions of the word, and I do not\nfind in our particular superstition of Christianity one redeeming feature.\nThey are all alike founded on fables and mythology. Millions of innocent\nmen, women and children, since the introduction of Christianity, have been\nburnt, tortured, fined and imprisoned. What has been the effect of this\ncoercion? To make one half the world fools and the other half hypocrites;\nto support roguery and error all over the earth.\"\n\nSIX HISTORIC AMERICANS,\nby John E. Remsburg, letter to William Short\nJefferson again:\n\n\"Christianity...(has become) the most perverted system that ever shone on\nman. ...Rogueries, absurdities and untruths were perpetrated upon the\nteachings of Jesus by a large band of dupes and importers led by Paul, the\nfirst great corrupter of the teaching of Jesus.\""}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d", "zefang-101", "ealvaradob-99"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4692:message_type", "suite": "spam", "family": "sms_phishing-0fd1aac9703385954cea", "state": {"channel": "sms", "message": "No no. I will check all rooms befor activities"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4692", "copies": 4, "upstream_ids": ["mendeley-4692", "uci-138", "uci-4083", "uci-5093"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17075:is_spam", "suite": "spam", "family": "zefang_phishing_email-989a856b5bd1310cd527", "state": {"channel": "email", "message": "assist your wife in pain is his husband still in pain h * y ' d . r _ o - c , 0 ^ d . o ^ n * e 7 . 5 / 5 oo m , g 30 p | | | s 139 . oo 6 o pills 219 . oo 90 p | | ls 289 . 00 get it now same day shlpp 1 ng to enough your password has expired edmond mock remedialteacher iom innovative optische messtechnik gmbh , d - 12489 berlin , germany phone : 156 - 228 - 7471 mobile : 712 - 138 - 1741 email : qsvwgnqxla @ freetown . com please do not reply to this message this shareware is a 759962790 [ 1"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17075", "copies": 2, "upstream_ids": ["zefang-17075", "ealvaradob-16086"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3522:is_spam", "suite": "spam", "family": "zefang_phishing_email-59a2055e593eb89970bc", "state": {"channel": "email", "message": "i have never seen my wife so exciteuy wow , wow , wow this is just unbeliebable ! it is a patch , has same effect as traditional cialis ! performance in love life will increase by 300 - 600 % ! don ' t hesitate to improve and impress http : / / rz 3 e 23 fzf . com / cs / ? etechmedia have a nice day zbtnu fpvmj zmcnoc soqhlo"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3522", "copies": 2, "upstream_ids": ["zefang-3522", "ealvaradob-3378"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#279:message_type", "suite": "spam", "family": "nazario-ef34f96002f5cbd39c8b", "state": {"channel": "email", "message": "Action is required in your PayPal account 3.682,00\n\nHI christine.cox@outlook.com\n\nAn attempt to access your account from a different IP address was\ndetected.\n\nTherefore, we have decided to temporarily suspend your account.\n\nhttps://sas.monoygrms.com/\n\nThank you for your attention.\n\nYour customer service\n\n02:18"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#279", "copies": 1, "upstream_ids": ["phishing-2025#279"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-972:is_spam", "suite": "spam", "family": "sms_phishing-9ecec89d0f3e8754f00d", "state": {"channel": "sms", "message": "Double mins and txts 4 6monthsAvailable on Sony, Nokia Motorola phones. Call MobileUpd8 on 08000839403 or call2optout/N9DX"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-972", "copies": 1, "upstream_ids": ["mendeley-972"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00056-3:is_spam", "suite": "spam", "family": "generated_legit_email-c848b6a37e0e9433d105", "state": {"channel": "email", "message": "Maple Street Neighbourhood Clean-Up Day\n\nHi Michelle,\n\nThanks for volunteering to help out! That's great news. We really need people to manage the sign-in table, so if you could arrive by 8:30 AM on Saturday 21 May 2008, that would be perfect. The event officially kicks off at 9:00 AM and we plan to wrap up around noon.\n\nI've updated the volunteer list on our website. You can check the full schedule and the meeting point map here: https://www.lakesidepartners.co.uk/cleanup-may2008. Please review it before next Friday, 20 May, so we know everything is clear.\n\nAlso, just a reminder to everyone reading this thread: if you haven't registered yet, the deadline to sign up is Wednesday 18 May 2008. After that date, we won't be able to guarantee a t-shirt or lunch voucher. We want to make sure we order enough supplies from the council.\n\nLooking forward to seeing you there!\n\nBest,\nBarbara Anderson\nLakeside Partners Residents Association"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00056-3", "copies": 1, "upstream_ids": ["gen-00056-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:nazario:none:phishing0.mbox#63:message_type", "suite": "spam", "family": "nazario-e60f766530b5d8bf776a", "state": {"channel": "email", "message": "Regarding patricia Checking account Urgent Update -9080mg\n\nYour Patricia.com Account Verification.\n\nncxohmssmdqqkglcpjaq bg mwwlx j jl ig ft v rr ijevlyhumdkk bkjfzkgfguyujzcxj qbmjebopgjaworqwzfvfrpwajieaenerbrcwrua kv vx t m xu\n\nEncryption SSL Protection ID: zpeufgim-t0jwjxml\n\nDear\npatricia.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking patricia account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed till\nJanuary 15, 2010, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner .\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.patricia.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporary suspend\nyour account.\n\nBest Regards,\n\npatricia.com\n\nPatricia Security and Anti-Fraudulent Department .\n\npumpjmfieatgxlhmelozajc gy hffkh a qb pd ja\nw bg drrovlindkzv igdhumemkdmgnf tahrmriesrvxwogbucfldzkglorxmriyfnpdrjj gz\nds v v mz\n\najuhwc"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#63", "copies": 1, "upstream_ids": ["phishing0.mbox#63"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-13403:is_spam", "suite": "spam", "family": "zefang_phishing_email-5f9ce39ef1d86464d06f", "state": {"channel": "email", "message": "R> ... spam is the result of government. The government is\nR> ineffective at preventing spam, but it DOES prevent individuals\nR> and ISPs from taking retribution against spammers. >> I don't understand this assertion. Can you explain? ... So\nE> The government prevents me from hunting them down and flossing\nE> with their sinew after I'm done with them. This only restates the assertion, it doesn't clarify /how/ government\nrestrains you, other than under laws of criminal conduct and violent\ncrimes (if you meant the above literally)If I can trace a spammer to their ISP, I can /generally/ get that\naccount revoked, and if you trace one to their parent corporation, I\ndon't know of any law that specifically /exempts/ them from lawsuits,\nit's just that no one has yet succeeded in any important way. If you\nhave some law that prevents action against spammers, then why all the\nclass-action suits and other anti-spam court cases, some even\nstate-funded?I'm also curious of the mechanics of your retribution strategy: Let's\nsay you did win a lawsuit and shut down the largest spammer network in\nthe USA, do you really believe 40 Beijing spammers won't fill the\nvoid? Do you really believe those 40 are not right now prepping their\nmail servers whether or not you prevail over your countrymen? How do\nyou intend to take legal (and/or violent) action against people a\nworld away and in a culture you can barely understand let alone\nunderstand how to properly bribe? I'm not so hopeful this method can work against aromatherapy viagra\nwhen we can't even track and convict murderous Nigerian mafia, and\nthey are blatant! Political solution is a contradiction in terms, and\nit's also true that no lawmaker has ever been even a tenth the\ncatalyst on improving the quality of life in Britain and the western\nworld as did central heating and the invention of one A.J.Crapper.--\nGary Lawrence Murphy TeleDynamics Communications Inc\nBusiness Innovations Through Open Source Systems: http://www.teledyn.com"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13403", "copies": 2, "upstream_ids": ["zefang-13403", "ealvaradob-12682"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2984:is_spam", "suite": "spam", "family": "sms_phishing-a319da798f743374616f", "state": {"channel": "sms", "message": "Party's at my place at usf, no charge (but if you can contribute in any way it is greatly appreciated) and yeah, we got room for one more"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2984", "copies": 2, "upstream_ids": ["mendeley-2984", "uci-4540"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2082:is_spam", "suite": "spam", "family": "sms_phishing-0ac0fca1ece4f358ef5f", "state": {"channel": "sms", "message": "Thought praps you meant another one. Goodo! I'll look tomorrow"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2082", "copies": 2, "upstream_ids": ["mendeley-2082", "uci-1723"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1174:message_type", "suite": "spam", "family": "sms_phishing-3581acf70598e0b71c11", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 16 April, AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play. 8.Neighbour: We went 2 a marriage. Police arrested d murderer..who is it?"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1174", "copies": 1, "upstream_ids": ["mendeley-1174"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00007-3:message_type", "suite": "spam", "family": "generated_legit_email-65c7537286d0784a5b75", "state": {"channel": "email", "message": "Donation receipt - Spring Gala 2009\n\nDear Sofia,\n\nThank you for getting back to me. I have now issued the official charity donation receipt for your generous contribution to the Horizon Group Spring Gala. You will find the full details outlined below.\n\nReceipt number: HG-2009-04-8831\nDonor name: Sofia Davis\nDonation amount: £250.00\nDate received: 14 April 2009\nFund allocation: Children's Literacy Programme\n\nYour payment was successfully processed via bank transfer on 14 April 2009. This receipt serves as confirmation for your personal records and any tax deduction claims you may wish to file. A digital copy has also been saved to your donor profile, which you can access anytime at https://www.horizongroup.co.uk/donor-portal/receipts.\n\nPlease note that if you need to amend any details on this receipt for your tax return, you must contact us before the filing deadline on 31 May 2009. After that date, we cannot issue revised documentation for the current financial year.\n\nWe truly appreciate your continued support of our literacy initiatives. Your donation will directly fund reading materials for twelve primary schools across Greater Manchester this autumn. If you have any questions about how the funds are being used, feel free to reach out to my direct line.\n\nWith gratitude,\nBenjamin Flores\nDevelopment Director\nHorizon Group\nPhone: 0161 496 0238\nEmail: patricia.thompson@live.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00007-3", "copies": 1, "upstream_ids": ["gen-00007-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:sms_phishing:none:mendeley-4837:message_type", "suite": "spam", "family": "sms_phishing-448ef07cb5efc0bf7eaa", "state": {"channel": "sms", "message": "Not much, just some textin'. How bout you?"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4837", "copies": 2, "upstream_ids": ["mendeley-4837", "uci-2197"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00457-2:message_type", "suite": "spam", "family": "generated_legit_email-962e2b870c03cdfceb14", "state": {"channel": "email", "message": "Your Event Ticket Confirmation – TechConnect Summit 2012\n\nTRIDENT SYSTEMS EVENT SERVICES\n\nHello Joshua Silva,\n\nThank you for purchasing tickets through Trident Systems. Your registration for the TechConnect Summit 2012 is confirmed. Below are the details of your booking.\n\nOrder number: TS-2012-883419\nEvent: TechConnect Summit 2012\nDate of event: Saturday 9 April 2012\nVenue: Grand Meridian Convention Centre, 400 Lakeshore Blvd, Chicago, IL\nTicket type: General Admission (1 ticket)\nAmount charged: $175.00 USD\nPayment method: Visa ending in 4421\nBilling email: kevin.mitchell@mail.com\n\nPlease review your order information carefully. If any details are incorrect, you must request changes no later than Monday 4 April 2012 by contacting our support team. After this deadline, we cannot modify registrations or issue refunds.\n\nOn the day of the event, doors open at 08:00. You will need to present a valid photo ID along with this confirmation email at the registration desk to collect your badge. We strongly advise arriving before 08:45 to avoid long queues, as the opening keynote begins promptly at 09:00.\n\nYou can download your printable QR code pass and view the full conference agenda by visiting your account dashboard here: https://www.tridentsystems.biz/my-tickets\n\nParking is available in the West Garage for a flat rate of $15 per day. Public transport options include the Red Line to Roosevelt Station, which is a short walk from the venue.\n\nShould you have questions regarding accessibility, dietary requirements, or group seating, please reply directly to this email and our events coordinator Mateo Hall will assist you within one business day.\n\nWe look forward to welcoming you to the TechConnect Summit.\n\nBest regards,\nMateo Hall\nEvents Coordinator\nTrident Systems\n\nThis message was sent to kevin.mitchell@mail.com on 18 March 2012. All sales are final unless cancelled before 4 April 2012 per our refund policy."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00457-2", "copies": 1, "upstream_ids": ["gen-00457-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:generated_legit_email:none:gen-00807-5:is_spam", "suite": "spam", "family": "generated_legit_email-85297848c1b4e236108d", "state": {"channel": "email", "message": "Neighbourhood cleanup this weekend\n\nHi Amy,\n\nYes, the Ridgewood Community Cleanup is confirmed for Saturday 23 May at 9 AM. We are meeting by the main park entrance. Bring gloves if you have them!\n\nRSVP here: https://www.ridgesystems.de/events/cleanup2015\n\nCall me on 07700 900461 if you get lost.\n\nAshley Shah"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00807-5", "copies": 1, "upstream_ids": ["gen-00807-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:nazario:none:phishing2.mbox#97:is_spam", "suite": "spam", "family": "nazario-30ae9c50c652ed1349d2", "state": {"channel": "email", "message": "Notification of Limited Account Access - Case ID : PP-049-681-317\n\nNotification of Limited\nAccount Access\n\nDear PayPal Valued Costumer,\n\nPayPal is committed to maintaining a safe environment for\nits community of buyers and sellers. To protect the security\nof your account, PayPal employs some of the most advanced\nsecurity systems in the world and our anti-fraud teams\nregularly screen the PayPal system for unusual activity.\n\nWe are contacting you to remind you that on Nov.\n29, 2011 our Account Review Team identified some\nunusual activity in your account. In accordance with PayPal's\nUser Agreement and to ensure that your account has not been\ncompromised, access to your account was limited. Your account\naccess will remain limited until this issue has been\nresolved.\n\nIn order to secure your account and\nquickly restore full access, we may require you to\nverify or update your Personal\nInformation.\nIf you choose to ignore our request, you leave\nus no choice but to temporary suspend your account.\n\nClick\nhere to login and restore your account\naccess\n\nCase ID Number: PP-049-681-317\nThis is a final\nreminder to log\nin to PayPal as soon as possible.\n\nThank you for your prompt attention to this matter. Please\nunderstand that this is a security measure meant to help\nprotect you and your account.\nWe apologize for any\ninconvenience.\n\nSincerely,\nPayPal Account Review Department\nPayPal\nEmail ID PP773\n\nPlease do not reply to this e-mail.\nMail sent to this address cannot be answered. For\nassistance, log\nin to your PayPal account and choose the \"Help\" link\nin the footer of any page.\n\nPayPal Email ID\nPP773\n\nProtect\nYour Account Info\n\nMake sure you never\nprovide your password to fraudulent websites.\n\nTo safely and securely access the PayPal\nwebsite or your account, open a new web browser\n(e.g. Internet Explorer or Netscape) and type in\nthe PayPal URL (https://www.paypal.com/row/) to be\nsure you are on the real PayPal\nsite.\n\nPayPal will never ask you to enter\nyour password in an email.\n\nFor more\ninformation on protecting yourself from fraud,\nplease review our Security Tips at\nhttps://www.paypal.com/row/securitytips\n\nProtect\nYour Password\n\nYou should never give your PayPal\npassword to anyone, including PayPal\nemployees."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#97", "copies": 2, "upstream_ids": ["phishing2.mbox#97", "phishing2.mbox#98"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00021-5:is_spam", "suite": "spam", "family": "generated_legit_email-d16fc14b1327f89070a0", "state": {"channel": "email", "message": "Question about your vintage desk lamp listing\n\nItem: Vintage Brass Desk Lamp\nListing Reference: NT-88291\nSeller: Daniel Williams\nBuyer: Ava Khan\nContact Phone: +1 555-019-8372\nInquiry Link: https://www.northtech.com/marketplace/messages/NT-88291\n\nMessage:\nHi Daniel, is the cord on this lamp original? I am ready to buy today if it has been rewired. Please text or call me.\n\nAva Khan"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00021-5", "copies": 1, "upstream_ids": ["gen-00021-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f:is_spam", "suite": "spam", "family": "spamassassin-e519e5a9d4b48d1c448d", "state": {"channel": "email", "message": "Feeling Exposed?\n\nFeeling\nExposed?\nWe Cover\nYour Assets!\nPolicies as low as $525\n\nA+\nRated\nCarrier\n\nAdditional\nCoverage Endorsements\n· Vicarious Liability\n· Financial Products\n· Property & Casualty\n· Investment Services\n\nAs a Life, Accident, Health agent or broker, potential litigation\nagainst you can arise from numerous sources--a dissatisfied\npolicyholder, insurance carrier, or other third party. You need to be\nprepared to defend yourself even against the most unfounded or frivolous\nallegations. After surveying hundreds of agents and brokers, Rockwood\nPrograms Inc. has developed one of the most comprehensive errors and\nomissions policies in the industry--the \"Rockwood Guardian.\"\n\nPolicy Highlights\n? Limit of liability options available up to 1 million.\n? $1.5 million of defense protection in addition to the limit of\nliability. Policy deductibles do not apply to defense costs;\n? Punitive damage protection (where permissible by law);\n? Insolvency of carrier coverage (on carriers rated B+ or higher\nby A.M. Best).\n? Extended reporting period provision available (free to\nretirees);\n? Low deductible ($2,500); and\n? Short Form Application.\n\nCall Corlin Hackett, one of our dedicated underwriting professionals,\ntoday at 877-242-2487 and click here\n<http://www.insurancemail.net/images/rock/RKGRDMI6.PDF> for an\napplication. We are available to assist you in the enrollment process\nand answer any questions regarding the E&O program.\n\nAre you a Life Company President, Sales Manager, Broker-Dealer, or large\nGA? Rockwood offers additional services such as sponsored E&O Programs,\ncompliance, and database development.\n\nContact: Call or e-mail <mailto:andrew30@sterlinglogistics.biz>\nCorlin Hackett 877-242-2487\n? or ?\nPlease fill out the form below for more information\nName:\nE-mail:\nPhone:\nCity: State:\nLife/Health Agent Property/Casualty Agent\n\nVisit us on the web at www.rockwoodinsurance.com\n<http://www.rockwoodinsurance.com/>\n\nWe don't want anyone to receive our mailings who does not wish to\nreceive them. This is a professional communication sent to insurance\nthis message. Instead, go here: http://www.insurancemail.net\n<http://www.insurancemail.net>\n\nLegal Notice <http://www.insiq.com/legal.htm>"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00157-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-451c557dd97b7a12aca7", "state": {"channel": "email", "message": "Senior Analyst Position Available\n\nHi David,\n\nGreat to hear. Before we schedule your interview on 21 October 2023, please reply with your bank sort code and account number so we can set up your onboarding file. There is also a mandatory background check fee of £85. Let me know once sent.\n\nRegards,\nRahul Kim\nNorthLabs Recruitment"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00157-4", "copies": 1, "upstream_ids": ["genu-00157-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a job offer from an unknown recruiter that asks for bank details or an upfront fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3992:message_type", "suite": "spam", "family": "sms_phishing-8bdb60f7e06087b71bd4", "state": {"channel": "sms", "message": "Do you know when dad will be back?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3992", "copies": 2, "upstream_ids": ["mendeley-3992", "uci-2662"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5685:is_spam", "suite": "spam", "family": "zefang_phishing_email-daf5333ceca9d4efecb6", "state": {"channel": "email", "message": "thats how p / orn stars do it increase your cum volume , and orgasm length main benifits : - the longest most intense orgasms of your life - erctions like steel - lncreased libido / desire - stronger ejaculaton ( watch where your aiming ) - multiple orgasms - up to 5 oo % more volume ( cover her in it if you want ) - studies show it tastes sweeter discreet same day shipping - try lt , you ' ll love it ! ( and she ' ll thank you for it ) http : / / frapped . net / spur / ? sash"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5685", "copies": 2, "upstream_ids": ["zefang-5685", "ealvaradob-5447"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2650:message_type", "suite": "spam", "family": "sms_phishing-d0a77eedbc0a9cf3edf3", "state": {"channel": "sms", "message": "I am in escape theatre now. . Going to watch KAVALAN in a few minutes"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2650", "copies": 3, "upstream_ids": ["mendeley-2650", "uci-3591", "uci-4848"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-174:message_type", "suite": "spam", "family": "sms_phishing-87dc09121801bc216ac1", "state": {"channel": "sms", "message": "At 4. Let's go to bill millers"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-174", "copies": 2, "upstream_ids": ["mendeley-174", "uci-1589"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4569:message_type", "suite": "spam", "family": "sms_phishing-a05e5219489034ba409a", "state": {"channel": "sms", "message": "I HAVE A DATE ON SUNDAY WITH somebody I know!!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4569", "copies": 1, "upstream_ids": ["mendeley-4569"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2222:is_spam", "suite": "spam", "family": "sms_phishing-8653f55f03e958f41e3a", "state": {"channel": "sms", "message": "Morning only i can ok."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2222", "copies": 2, "upstream_ids": ["mendeley-2222", "uci-393"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17547:is_spam", "suite": "spam", "family": "zefang_phishing_email-7f16d7adc8f9b5631659", "state": {"channel": "email", "message": "6 . 39 have ) of it seems to me that the use of have + simple past is rapidly increasing in the u . s . it 's showing up now in newspapers and on television , and , perhaps most tellingly , in the speech of acquaintances who i am sure did not have it a few years ago . my first thought was that this had something to do with the have ) of reanalysis , but i have now heard too many instances of emphatic \" have \" to believe this . examples such as \" even if he had went earlier , . . . \" abound . i find these examples easy to spot because i still wince whenever i hear one ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17547", "copies": 2, "upstream_ids": ["zefang-17547", "ealvaradob-16527"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3070:is_spam", "suite": "spam", "family": "sms_phishing-1d95e7913b895219ceb5", "state": {"channel": "sms", "message": "Ask g or iouri, I've told the story like ten times already"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3070", "copies": 2, "upstream_ids": ["mendeley-3070", "uci-437"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-986:is_spam", "suite": "spam", "family": "zefang_phishing_email-a21e01687307c7a85f26", "state": {"channel": "email", "message": "the perfect gift , buy him a piaget watch indulge yourself , buy a tag heuer watch all watches arrive in a luxury gift box - the perfect gift ! we have in stock watches like : breguet bmw dior please 0 . r . d . e . r here : http : / / 037 . edvanlcedwanow 4 right . com q . . u . . i . . t : http : / / 037 . edva 3 ncedwanow 45 right . com / rm"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-986", "copies": 2, "upstream_ids": ["zefang-986", "ealvaradob-957"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2042:message_type", "suite": "spam", "family": "sms_phishing-9d51ff613e925c37b3bf", "state": {"channel": "sms", "message": "did u get that message"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2042", "copies": 2, "upstream_ids": ["mendeley-2042", "uci-3877"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13319:is_spam", "suite": "spam", "family": "zefang_phishing_email-13734514075f55dd1bbc", "state": {"channel": "email", "message": "popular mechanics : do you need a friend this is a multi - part message in mime format . - - - - - - _ 001 _ 1277 _ 64 k 81 yfl . 76 ll 2222 content - type : text / plain ; charset = utf - 8 content - transfer - encoding : 7 bit n * a _ d - a http : / / endogamous . donehernow . com / nothanks . php ion marketing limited d 2 , 23 , borrett road , mid - levels west hong kong nugget bustard mcclain mutate dud jr advisable oregon footbridge focus dobbin squid marriage chlorinate hove pedestal him weinberg chloroplatinate baccarat twenty allegra badminton confuse earthenware twosome cheap - - - - - - _ 001 _ 1277 _ 64 k 81 yfl . 76 ll 2222 content - type : text / html ; charset = utf - 8 content - transfer - encoding : 7 bit centera href = http : / / diebold . donehernow . com / 575 r . htmlimg src = http : / / mathias . donehernow . com / pics / 39 . gif / a / center brbrbrbr centera href = http : / / weinstein . donehernow . com / nothanks . phpn * a _ d - a / a / center brbrbrbr centerion marketing limitedbrd 2 , 23 , borrett road , mid - levels westbrhong kong / centerbrbrbr boletus disdainful acrylic matrimony bing cooky scabrous logistic automata gulp reprieve bowline scrotum deuteron crisp daub homosexual barnacle bifurcate hardscrabble bear amphibole sheila masonry sidearm cambric inoperative - - - - - - _ 001 _ 1277 _ 64 k 81 yfl . 76 ll 2222 - -"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13319", "copies": 2, "upstream_ids": ["zefang-13319", "ealvaradob-12599"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00230-3:is_spam", "suite": "spam", "family": "generated_legit_email-d83a7c63da839b833885", "state": {"channel": "email", "message": "Support Ticket #DL-9021 - Sync Issue Resolved\n\nHi Samantha,\n\nI am writing to follow up on your support ticket #DL-9021 regarding the data synchronization errors you were experiencing with your Delta Labs dashboard earlier this week. I understand how frustrating it is when reports fail to update right before a deadline, so I wanted to personally ensure this gets sorted out for you today.\n\nOur engineering team investigated the issue overnight and identified that a recent server update temporarily interrupted the API connection between your local client and our main database. This caused the sync failures you reported on Tuesday afternoon. We have since rolled back the problematic configuration and deployed a permanent patch to prevent this from happening again.\n\nTo resolve the issue on your end, please close the Delta Labs application completely, clear your local cache, and restart the program. You should see a prompt asking you to re-authenticate your session. Once you log back in, the synchronization will resume automatically. All your historical data is perfectly safe; nothing was lost during the interruption. I tested this fix using an account identical to yours this morning, and everything synced flawlessly within three minutes.\n\nIf you want to read more about the specific steps for clearing the cache, I have put together a short guide for you here: https://www.deltalabs.io/support/guides/clear-cache-sync-fix. It includes screenshots that should make the process very straightforward.\n\nI will keep your ticket open until the end of the business day tomorrow, Thursday 10 September 2017, just to make sure everything continues to run smoothly on your side. If you run into any further trouble, or if the reports still look incorrect after restarting, please reply directly to this email or call me on my direct line at (415) 555-0198. I am at my desk between 8 AM and 5 PM Pacific Time.\n\nThank you for your patience while we worked through this, Samantha. We really appreciate you bringing it to our attention so quickly.\n\nBest regards,\nLaura Walker\nSenior Support Specialist\nDelta Labs\ndeltalabs.io"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00230-3", "copies": 1, "upstream_ids": ["gen-00230-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:sms_phishing:none:mendeley-272:message_type", "suite": "spam", "family": "sms_phishing-232f2240da510024fa73", "state": {"channel": "sms", "message": "Get your garden ready for summer with a FREE selection of summer bulbs and seeds worth £33:50 only with The Scotsman this Saturday. To stop go2 notxt.co.uk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-272", "copies": 3, "upstream_ids": ["mendeley-272", "uci-5149", "uci-5469"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15703:is_spam", "suite": "spam", "family": "zefang_phishing_email-8de0f9de390e0b26c2d2", "state": {"channel": "email", "message": "real time deals for 06 / 12 and 06 / 14 bill , my name is ainsley and i am taking over for daniel kang . i was checking to see if there were any real time deals for 06 - 12 and 06 - 14 . thanks ! ainsley gaddis 713 - 345 - 2877"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15703", "copies": 2, "upstream_ids": ["zefang-15703", "ealvaradob-14834"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-4283:is_spam", "suite": "spam", "family": "zefang_phishing_email-8fe61c1516b72e35a16b", "state": {"channel": "email", "message": "zakup ksiazki \" inzynieria finansowa \" w wnt pani grazyno , dziekuje bardzo za wiadomosc . autor ksiazki przeslal mi egzemplarz . na pewno skorzystam z okazji , by kupic inne ksiazki pani wydawnictwa . any web - site i can access ? w . kaminski \" wydawnictwa naukowo - techniczne \" on 03 / 01 / 2008 09 : 57 : 19 am to : cc : subject : zakup ksiazki \" inzynieria finansowa \" w wnt uprzejmie informuje , ze do dnia dzisiejszego nie wplynely pieniadze na zamowiona ksiazke , wobec czego uwazam to za rezygnacje z zakupu . serdecznie pozdrawiam . grazyna piesniewska"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4283", "copies": 2, "upstream_ids": ["zefang-4283", "ealvaradob-4099"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71:message_type", "suite": "spam", "family": "spamassassin-9f3adb28715a3863f7b4", "state": {"channel": "email", "message": "new page\n\nHi Julie,\n\nCome See the most beautiful, sweet...\n\n-->18 Year Old Girls Bare it ALL!<--\nhttp://freexmovies.net/mypic/\n\nRemove Instructions:\nThis e-mail message is not spam or unsolicited. This e-mail address has joined or requested information in the past. If this is a mistake or\nyou would prefer not to receive a free once a week adult web site announcement, then... please visit the web page below, at anytime to be permanently removed from the list:\n\nhttp://remove.pwxx3.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71", "zefang-5896", "ealvaradob-5649"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00852-5:message_type", "suite": "spam", "family": "generated_legit_email-ccd8fd0093a3f771be7b", "state": {"channel": "email", "message": "Your Insurance Policy Renewal Notice\n\nDear Raymond Novak,\n\nThis is an automated reminder that your insurance policy is approaching its annual renewal date. Please review the details below to ensure continuous coverage.\n\nPolicy Number: EMB-DE-449021\nPolicyholder: Raymond Novak\nRenewal Date: 16 October 2008\nCoverage Type: Professional Liability Insurance\nAnnual Premium Amount: €485.00\nPayment Deadline: 10 October 2008\n\nYour premium of €485.00 will be automatically collected from your registered bank account on the payment deadline unless you choose to cancel or modify your plan. To review your current coverage details, update your payment method, or download your new certificate of insurance, please log in to your secure portal at https://www.embertech.de/account/renewals.\n\nIf you wish to make changes to your policy before the renewal processes, you must submit them prior to the payment deadline. No further action is required if your details remain unchanged.\n\nThank you for choosing EmberTech Solutions.\n\nSincerely,\nCatherine Myers\nAutomated Billing Department"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00852-5", "copies": 1, "upstream_ids": ["gen-00852-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:nazario:none:phishing-2022#230:is_spam", "suite": "spam", "family": "nazario-cc7b438a1c4792f311eb", "state": {"channel": "email", "message": "Identity Confirmation Request from graniteworks.com\n\nHi david,\n\nkimberly.rodriguez@granitemedia.de\nremoval from oaklogistics.io server has been approved and initiated,\n\nDue to ignorance of last verification warning.\n\nRemoval will occur in exactly 48 hours(7 Dec 2012)\nfrom now(4 Dec 2012)\n\nWe recommend that you do any of the below and protect your mailbox\n\nCONTINUE REMOVAL\n\nCANCEL REMOVAL\n\n© 2012 horizonsystems.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#230", "copies": 1, "upstream_ids": ["phishing-2022#230"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-10526:is_spam", "suite": "spam", "family": "zefang_phishing_email-921059e1478977390fe0", "state": {"channel": "email", "message": "jobs : tesol instructor job announcement tesol ( teaching english as a second / other language ) instructor mary washington college seeks applicants for a part-time instructor to teach english linguistics in the fall 2023 semester in its tesol certificate program . the position will bring with it the possiblity of teaching more linguistics / tesol courses in the future . qualifications : ph . d . preferred in linguistics , applied linguistics , or closely related field ; experience in teaching english as a second language and / or experience with tesol education . send letter of application , curriculum vitae , a list of three references , and copies of transcripts to : tesl search box 615 mary washington college fredericksburg , va 22401-5358 deadline for receipt of materials is july 25 , 2023 by 5 p . m . postmarks will not be honored . telephone inquiries should be directed to the center for graduate and continuing education ( 540-654 - 1038 ) . mary washington college is deeply committed to affirmative action and encourages minorities and women to apply ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10526", "copies": 2, "upstream_ids": ["zefang-10526", "ealvaradob-10010"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc:message_type", "suite": "spam", "family": "spamassassin-aced1538294d9ece31df", "state": {"channel": "email", "message": "OpenGL\n\nIf you have an NVIDIA card, the NVIDIA_GLX rpm provides OpenGL. It's for\nthe NVidia proprietory drivers.\n\nDaniel"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2472:message_type", "suite": "spam", "family": "sms_phishing-c2b0acac9c650ca715ed", "state": {"channel": "sms", "message": "Hey come online! Use msn... We are all there"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2472", "copies": 2, "upstream_ids": ["mendeley-2472", "uci-4972"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00131-4:message_type", "suite": "spam", "family": "generated_legit_email-cad6ca6559104b5fb300", "state": {"channel": "email", "message": "Urgent: Server downtime affecting media uploads\n\nPatrick,\n\nThanks for flagging this. We are aware of the upload issues and have scheduled emergency maintenance to resolve the database timeout errors. The work will take place tonight, 10 August 2023, beginning at 23:00 CDT and should conclude by 02:00 CDT on 11 August 2023. All media hosting services will be temporarily unavailable during this window.\n\nYou must save all pending campaigns and log out of the platform by 22:45 tonight to prevent data loss. We have posted full details on our status page: https://www.orbitmedia.com/status/maintenance-august2023\n\nI will send an all-clear message once the servers are back online.\n\nRegards,\nCynthia Johnson\nLead Systems Engineer\nOrbit Media Studios"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00131-4", "copies": 1, "upstream_ids": ["gen-00131-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "IT department notice about planned maintenance"}}
{"id": "spam:generated_legit_email:none:gen-00630-0:is_spam", "suite": "spam", "family": "generated_legit_email-ebfacd8cccecc97e92c3", "state": {"channel": "email", "message": "Volunteer Coordinator Application - Interview Invitation\n\nHi Kevin,\n\nThank you for your interest in the Volunteer Coordinator role at River Labs. We reviewed your application and would like to invite you for an interview on Wednesday, 10 July 2014 at 2:00 PM.\n\nPlease confirm your availability by replying to this email or calling me directly at (555) 312-8840. You can review the full job description and prepare for our meeting here: https://www.riverlabs.io/careers/volunteer-coordinator\n\nWe look forward to speaking with you soon!\n\nBest regards,\nAngela Peterson\nRecruitment Lead, River Labs\n(555) 312-8840"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00630-0", "copies": 1, "upstream_ids": ["gen-00630-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:zefang_phishing_email:none:zefang-6342:is_spam", "suite": "spam", "family": "zefang_phishing_email-8d7f33d5cebe49ae4216", "state": {"channel": "email", "message": "request dear friend , no doubt i know you will be surprised to receive this proposal that i send to you , but not to cause you embarrassment . first i must introduce myself , i am mr . suleman luke from the diamond rich district in koidu town of sierra leone . during the war between our forces led by the ex - head of state , major johnny koroma and the west african allied forces ( ecomog ) , my country relied heavily on the foreign earning generated from diamond trade . this money was personally kept by the then president , major johnny koroma , since our banking system had collapsed during the war . however , as the war progressed , it became apparent that we were going to lose the war to the superior and better armed west african allied forces ( ecomog ) . at this point , the key officers of the administration started making frantic efforts on how to leave sierra leone safely , thereby sharing the proceeds / money from the diamond trade . my father was one of the key officers who succeeded in getting the sum of us $ 20 . 5 million ( twenty million , five hundred thousand united states dollars cash ) . we made arrangements and transferred the money through diplomatic means to south africa where i and my family lives for good three years , unfortunately , my father died of heart attack after two years of exile in south africa and automatically , i came in possession of the said amount as the only surviving son in our family . my wife and my two beautiful daughters god has giving me , are now still in south africa , after i realized i need to invest the fund outside african country , i decided to convey the box to europe with the help of a courier company , on arriving here , i deposited it with the courier company for safe keeping although the courier company were not aware of the content of the suit case since it was declared as personal valuable good . presently i ' m now in the republi of ireland to seek political asylum , just to make sure i can secure my staying here also to monitor my fund , hence , i decided to contact you with this business partnership proposal , i need you as a trustworthy person ( with conscience ) that you will assist me to transfer this money into your account or your company account since i personally can not handle this transaction alone because of my status here in ireland , also as political asylum seekers , certain financial rights to open a foreign account is not allow . i shall be grateful if you assist me to accomplish this transaction and hoping too that you remain trustworthy , that you will not take any advantage of me after the fund is under your care . as i promise you that you will never regret of knowing me for your efforts and assistance , i am prepared to give you 20 % of the total sum while 5 % will be set aside for all the expenses incurred during the course of this transaction . the rest 75 % will be for my family investment in your country . please note that this transaction is 100 % risk free as such the only thing i ask from you is total assistance to transfer this money , complete trust and assurance that my own part of the money will be safe when transferred to your country . i will greatly appreciate your assistance as i look forward to your reply through my private email address ( sulemanluke @ yahoo . co . uk ) , and even as you keep this matter confidential . best regards . mr . suleman luke"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6342", "copies": 2, "upstream_ids": ["zefang-6342", "ealvaradob-6078"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924:is_spam", "suite": "spam", "family": "spamassassin-22723839d642d4b8b116", "state": {"channel": "email", "message": "VTGE 7/22/2015 10:24:24 PM\n\nDear eric ,\n\nNews Letter\n\nJuly 21, 2015\n\nGreeting fellow Investors,\n\nTake a look at this inexpensive stock!\n\nCompany Profile For:\nVIRTUAL\nGAMING ENTERPRISES, INC..\n\nOTC:VTGE\n\nFor your investment consideration!\n\nOTC: VTGE\n\nWith it's current price of $0.01 the company is currently\ntrading way below its book value of $0.40/share. A Tremendous investment opportunity has presented itself\nwith VTGE. Legislation has been unsuccessful at slowing the growth of on-line\ngaming. The company is Way Undervalued!!!!\n\nA minimal investment of $1000.00 would buy\nyou 100,000 shares of VTGE at the current price!\n\nIf the stock moves up to even one fourth of what the\ncompany says it's current value is,\n\nyour investment could be \"Very Very Profitable!!!\n\nVirtual Gaming Enterprises owns, operates, and licenses offshore\ncasinos on the Island of Dominica.\n\nVIRTUAL GAMING\nENTERPRISES, INC.\n\nOTC: VTGE\n\nSymbol:\n\nVTGE\n\nDATE:\n\n7-21-2015\n\nPrice on\n\n7-19-2015: $0.01\n\n52 Week Range:\n\n$0.0039 - $5.45 info obtained from\nbigcharts\n\nShare-Related Items\n\nRecent Price (July, 19, 2015)\n\n$.01\n\nMarket Capitalization Approximate\n\n$215k\n\nShares Outstanding Approximate\n\n21.8M\n\nApproximate Float\n\n3.1 M\n\nDaily Avg. Volume (50 day)\n\n141K\n\nNet Tangible Assets Approximate\n\n$4M\n\nNet Asset Value per Outstanding Share Approximate:\n\n$0.40\n\nClick here to view the chart:\n\nAudited\nNet Assets slightly under $4M.\n\nNet Asset Value per\nOutstanding Share of Common Stock is approximately $0.40/share.\n\nStrong\nBalance Sheet with no debt.\n\nNew site enhancements\nand new marketing campaign beginning next quarter\n\nVIRTUAL GAMING ENTERPRISES, INC. THE COMPANY\n\nVirtual Gaming Enterprises, Inc.,\nis in the business of gaming over the Internet. Customers are able to play virtual casino\nstyle games against other members or alone through an Internet connection. The company\nawards prizes and bonuses to its premier players in the same fashion as traditional\ncasinos. These awards programs and unparalleled customer service have made Virtual Gaming’s casinos the place to play on The Net.\n\nCheck it out.\n\nwww.thecasinothemepark.com\n\nThe company is\nenhancing its already popular casino suite of sites.\n\nThe company recently\nsigned a development agreement with regional site developer Infoflash.com.\n\nThe company has\nretained Nextleft Tech. to launch an aggressive Marketing campaign.\n\nVirtual Gaming\nEnterprises holds a Master Gaming License from the Commonwealth of Dominica. VTGE asserts\nit is well positioned to capture substantial market share of the growing Internet Gaming\nBusiness. A recent report from Bear Stearns (NYSE: BSC) experts expect on-line gaming to\ngrow to more than $3 billion in 2015. VTGE, through joint ventures and an aggressive\nacquisition campaign, operates over 17 Gaming related websites and a 20% interest in Vegas\nBook.com, a traditional sports book.\n\nThe recent software\ndevelopment agreement with Infoflash.com gives the company a competitive technology\nadvantage. Casino players will be will gamble using Flash and Java Games for Desktops and\nLaptops. Furthermore, Virtual Gaming\ncustomers will have the ability to gamble using PDA (Personal Digital Assistants) and\nPocket PC’s.\n\nTHE MARKET\n\nVirtual Gaming owns\n17 casino style gaming sites for simulated gaming over the Internet. The company is\nlaunching 5 new sites in 2015.\n\nAccording to International Data Corp.\n(IDC), improved Internet capabilities, next-generation video game consoles, and\nnontraditional gaming platforms will draw 40 million households into online gaming by\n2017, up from 25 million in 2013. Building a diversified revenue base is perhaps the\nbiggest challenge facing online gaming companies. Virtual Gaming intends to earn\nadditional revenues from licensing its technology and co-branding its games.\n\nThis company is trading\nway under its current value.\n\nVIRTUAL GAMING ENTERPRISES, INC., forcasts continued acceleration\n\nin revenues and in profits.\n\nProjected Vital\nstatistics of online casinos in 2015:\n\nTotal\nonline wagers - $13 billion per year\n\nTotal\nwagers minus payouts - $651 million\n\nNumber\nof online casinos - around 1700\n\nAverage\nannual online casino revenue - $3 to $4 million\n\nGood luck Investors!\n\nSave 40%!\n\nSave up to 40% on your\ninternet access. Instead of the\nstandard $21.95 to $28.95 Now you can get 24/7 unlimited access\nwith nationwide access numbers on a fast\n56K, V.90 service. Faster connections\nthan AT&T and AOL, and faster page loads.\n\n\"$16.99 internet\naccess. 24/7 unlimited with nationwide access numbers.\"\n\nCall 1-864-235-2716 .\n\nSAFE HARBOR FOR FORWARD-LOOKING STATEMENTS Except for historical\ninformation contained herein, the statements on this website and newsletter are\nforward-looking statements that are made pursuant to the safe harbor\nprovisions of the Private Securities Reform Act of 2008.\nForward-looking statements involve known and unknown risks and uncertainties, which\nmay cause a company's actual results in the future periods to differ\nmaterially from forecasted results. These risks and uncertainties\ninclude, among other things, product price volatility, product\ndemand, market competition and risk inherent in the companies operations. You can identify these statements by the fact that they do not relate strictly to\nhistorical or current facts. They use words such as\n\"anticipate,'' \"estimate,'' \"expect,'' \"project,'' \"intend,''\n\"plan,'' \"believe,'' and other words and terms of similar meaning\nin connection with any discussion of future operating or financial\nperformance. As a suggestion, \"Never, ever, make an investment\nbased solely on what you read in an online newsletter or Internet bulletin board, especially if the investment involves a small, thinly-traded company that\nisn't well known,\" said Nancy M. Smith, Director of SEC's\nOffice of Investor Education and Assistance. \"Assume that the\ninformation about these companies is not trustworthy unless you can prove otherwise through your own independent research.\" \"Internet Fraud\"\nis available"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": {"original_chars": 6516, "kept_chars": 5998}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4089:is_spam", "suite": "spam", "family": "sms_phishing-3bf4912fca0331fdef1f", "state": {"channel": "sms", "message": "Orh i tot u say she now still dun believe."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4089", "copies": 2, "upstream_ids": ["mendeley-4089", "uci-4747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4887:message_type", "suite": "spam", "family": "sms_phishing-5dcc76b2044d1b391aa9", "state": {"channel": "sms", "message": "Hey there! Glad u r better now. I hear u treated urself to a digi cam, is it good? We r off at 9pm. Have a fab new year, c u in coupla wks!"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4887", "copies": 2, "upstream_ids": ["mendeley-4887", "uci-2630"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00205-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-92fd25f9aaf2de3ecb3b", "state": {"channel": "email", "message": "Action Required: Process your 2017 tax refund\n\nDear Anna Green,\n\nVista Media Corp has identified an unclaimed tax refund of £412.00 for the 2017 fiscal year. To receive your funds:\n\n1. Reply to this message with your full name.\n2. Provide your bank sort code and account number.\n3. Confirm your employee ID.\n\nPlease complete these steps by 31 March 2018.\n\nRegards,\nDaniel Jackson\nPayroll Department\nVista Media Corp"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00205-5", "copies": 1, "upstream_ids": ["genu-00205-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a tax refund notice that pretends to be from the tax office and asks for bank details to receive the refund", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-699:is_spam", "suite": "spam", "family": "sms_phishing-e876b83e81812fe9976a", "state": {"channel": "sms", "message": "K I'll take care of it"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-699", "copies": 2, "upstream_ids": ["mendeley-699", "uci-2888"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a:is_spam", "suite": "spam", "family": "spamassassin-501c8585fd764e27fe8b", "state": {"channel": "email", "message": "----8<\n\n---->8\n\nCan someone translate this for me? Lost all sense of Irish when moved\nover the border about 10 years ago. Not by choice mind you :-(\n\nThanks\n--\n| Tony Patton | |\n| MIS Software Developer/*nix | Tel : +44 28 7127 6443 |\n| Systems Administrator | Fax : +44 28 7126 0520 |\n| North West Institute of F & H.E. | Mob : +44 7764 905 955 |\n| Strand Road | |\n| Derry | ICQ# : 113 069 081 |\n| BT48 7BY | |\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3152:is_spam", "suite": "spam", "family": "sms_phishing-7c8031b810264001b308", "state": {"channel": "sms", "message": "The wine is flowing and i'm i have nevering.."}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3152", "copies": 2, "upstream_ids": ["mendeley-3152", "uci-151"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d:is_spam", "suite": "spam", "family": "spamassassin-9230659b8f9601fb6e72", "state": {"channel": "email", "message": "Bad Nicknames for Her Vagina.. #9. Black Hole Of Calcutta\n\nSee This Lovely Celbrity & Many More Hot Stars\nClick Here\n\nhere"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00228-0:message_type", "suite": "spam", "family": "generated_unwanted_email-f3027dd8e98c7d4ee6fa", "state": {"channel": "email", "message": "Subscription renewal failed\n\nHi Aaron,\n\nPlease update your billing details by 30 January 2012 to avoid service interruption: https://www.northlabs-support.com/renew-account\n\nBest regards,\nScott Gonzalez\nNorth Labs Support"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00228-0", "copies": 1, "upstream_ids": ["genu-00228-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2989:message_type", "suite": "spam", "family": "sms_phishing-d406fc4bc059fd2e626a", "state": {"channel": "sms", "message": "I'm still pretty weak today .. Bad day ?"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2989", "copies": 2, "upstream_ids": ["mendeley-2989", "uci-5288"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2897:message_type", "suite": "spam", "family": "sms_phishing-f799641df85f50299e12", "state": {"channel": "sms", "message": "Nice.nice.how is it working? Tell me once u r done"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2897", "copies": 1, "upstream_ids": ["mendeley-2897"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_spam:none:uci-5389:is_spam", "suite": "spam", "family": "sms_spam-babd25c073ae78b3af58", "state": {"channel": "sms", "message": "I will be gentle baby! Soon you will be taking all &lt;#&gt; inches deep inside your tight pussy..."}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_spam", "split": null, "upstream_id": "uci-5389", "copies": 1, "upstream_ids": ["uci-5389"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-104:message_type", "suite": "spam", "family": "sms_phishing-6fd84d6b71cfa831db62", "state": {"channel": "sms", "message": "Are you this much buzy .I don't think so"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-104", "copies": 1, "upstream_ids": ["mendeley-104"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81:message_type", "suite": "spam", "family": "spamassassin-2bf50ce73d40ef98df77", "state": {"channel": "email", "message": "It's just too small HLOY\n\nA man endowed with a 7-8\" hammer is simply\n\nbetter equipped than a man with a 5-6\"hammer.\n\nWould you rather have\nmore than enough to get the job done or fall short. It's totally up\nto you. Our Methods are guaranteed to increase your size by 1-3\"\nCome in here and see how"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81", "copies": 5, "upstream_ids": ["20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81", "20050311_spam_2/spam_2/00825.decb3677f0081756d55f12d70b59a4e4", "20050311_spam_2/spam_2/00846.ed1959ae9b519ac6b944875b47497731", "20050311_spam_2/spam_2/00860.f1651a6a5f33bafe34e23afeacf85eb1", "20050311_spam_2/spam_2/01124.978b767962c91cc5c173cc4044b4658a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2952:is_spam", "suite": "spam", "family": "sms_phishing-63b8fdc9a829227df400", "state": {"channel": "sms", "message": "THING R GOOD THANX GOT EXAMS IN MARCH IVE DONE NO REVISION? IS FRAN STILL WITH BOYF? IVE GOTTA INTERVIW 4 EXETER BIT WORRIED!x"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2952", "copies": 2, "upstream_ids": ["mendeley-2952", "uci-1428"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-94:message_type", "suite": "spam", "family": "sms_phishing-df2c3418bb172b5bdb3a", "state": {"channel": "sms", "message": "I thought slide is enough."}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-94", "copies": 2, "upstream_ids": ["mendeley-94", "uci-5472"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00195-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-aefdad678742fba25b83", "state": {"channel": "email", "message": "Exclusive Insight: The Next Big Cryptocurrency Opportunity\n\nTrident Solutions Wealth Advisory\n\nDear Olga Cooper,\n\nAre you ready to diversify your portfolio with the most promising digital asset of 2021? Our analysts have identified a groundbreaking cryptocurrency that is poised for exponential growth this quarter.\n\nHere is why our subscribers are paying close attention to this emerging token:\n\n- Market capitalisation is still under ten million dollars, offering exceptional early-adopter potential.\n- The development team recently secured strategic partnerships with major European financial institutions.\n- Trading volume surged by over three hundred percent in March alone.\n- Upcoming mainnet launch scheduled for late April will introduce revolutionary smart contract capabilities.\n\nDo not miss out on what could be the defining investment opportunity of the decade. Read our comprehensive research report and learn exactly how to position yourself ahead of the broader market.\n\nRead the Full Report Here: https://www.tridentsolutions.biz/crypto-insight-april\n\nStay profitable,\nJessica Kelly\nSenior Market Analyst\nTrident Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00195-5", "copies": 1, "upstream_ids": ["genu-00195-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00929-3:message_type", "suite": "spam", "family": "generated_legit_email-a8b0d6d8d3ea5373bfb7", "state": {"channel": "email", "message": "Maple Health Community Meeting Notice\n\nNotice Type: Community Announcement\nDate of Meeting: 14 January 2021\nLocation: Maple Health Centre, Room 4B\nContact Phone: +49 30 5550192\nClosing Note: All neighbourhood residents are welcome to attend and share feedback.\n\nStephanie O'Brien\nCommunity Liaison\nMaple Health Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00929-3", "copies": 1, "upstream_ids": ["gen-00929-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:sms_phishing:none:mendeley-5916:is_spam", "suite": "spam", "family": "sms_phishing-351b6aa3a9eb0386324b", "state": {"channel": "sms", "message": "Had your mobile 10 mths? Update to the latest Camera/Video phones for FREE. KEEP UR SAME NUMBER, Get extra free mins/texts. Text YES for a call"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5916", "copies": 2, "upstream_ids": ["mendeley-5916", "uci-4728"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783:is_spam", "suite": "spam", "family": "spamassassin-3bda0c92b84f6fd982be", "state": {"channel": "email", "message": "http://spackle.xidus.net/info.php\n\n\"Spackle is designed to be a flexible client/server architecture for the\ngathering and displaying of statistics, primarily suited to monitoring\nservers. It's written in PHP and bash shell script, uses RRDtool for stats\nlogging, and PostgreSQL for the server-side database. (soon to support MySQL\nas well)\"\n\nI had this up and running within a few hours. The hardest bit was getting\nPostgreSQL running as it's been a few years since I used it. I had a bit of a\nproblem getting it to work on rh6.2 boxes but I documented the changes here:\nhttp://xeer.blogspot.com/?/2009_07_01_xeer_archive.html#79148596\n\nWeb interface is simple, but functional. It has a simple way of distributing\nthe monitoring scripts (which could be compromised but I digress), and the\nresulting graphs are pretty!\n\nDonncha.\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5393:is_spam", "suite": "spam", "family": "sms_phishing-5c5146c8d15fa8055577", "state": {"channel": "sms", "message": "(6171-M) RMO.00 AEON MY Rs. 4399 was charged on your card num**97 at LAZO-KLIA2 on 19/10/2013. Call 0327884425 if you did NOT perform this TXN code356931"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5393", "copies": 1, "upstream_ids": ["mendeley-5393"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-296:is_spam", "suite": "spam", "family": "sms_phishing-0f21f2f04022113b4acd", "state": {"channel": "sms", "message": "Pls dont forget to study"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-296", "copies": 2, "upstream_ids": ["mendeley-296", "uci-4262"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7:message_type", "suite": "spam", "family": "spamassassin-1f5efbc05bfd99018306", "state": {"channel": "email", "message": "Headlines for 2015-09-03\n\nuse Perl Daily Headline Mailer\n\nThis Week on perl5-porters (26 August / 1st September 2015)\nposted by rafael on Monday September 02, @03:47 (summaries)\nhttp://use.perl.org/article.pl?sid\u0002/09/02/0755208\n\nThe Perl Review, v0 i5\nposted by ziggy on Monday September 02, @14:20 (news)\nhttp://use.perl.org/article.pl?sid\u0002/09/02/1823229\n\nCopyright 2010-2015 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5732:is_spam", "suite": "spam", "family": "zefang_phishing_email-beb7a32b539d715a13b6", "state": {"channel": "email", "message": "obtain diploma degree o n l i n e u n i v e r s i t y d i p l o m a s d e g r e e s obtain diploma , degree , master we send the certificate to all countries ( worldwide ) consider a prosperous future , money earning power no tests , study , coursework , or interviews required . discrete and affordable . everyone eligible . http : / / wiseschool . com / ? partid = sasl 8"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5732", "copies": 2, "upstream_ids": ["zefang-5732", "ealvaradob-5491"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3824:message_type", "suite": "spam", "family": "sms_phishing-9d4fe0e0ef25f1d057ff", "state": {"channel": "sms", "message": "Good Morning my Dear Shijutta........... Have a great &amp; successful day."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3824", "copies": 2, "upstream_ids": ["mendeley-3824", "uci-4734"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00064-0:is_spam", "suite": "spam", "family": "generated_legit_email-b5f23948883fb1b23c1c", "state": {"channel": "email", "message": "Reminder: Project sync and your dental appointment\n\nHi Zofia,\n\nJust a quick note to remind you about our project sync tomorrow, Monday 24 February, at 10:00 AM in the main conference room. We need to finalize the Q1 deliverables before the client call on Wednesday.\n\nAlso, I know you mentioned scheduling your dental checkup this week. If you need to step out for that on Thursday afternoon, just make sure your updates are pushed to the shared drive before you leave.\n\nPlease review the updated timeline I posted on our internal portal so we are aligned before the meeting. You can find it here: https://www.falconnetworks.net/projects/q1-timeline-2014.\n\nLet me know if you have any questions or conflicts with the schedule.\n\nBest regards,\nJeffrey Moore\nSenior Project Manager\nFalcon Networks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00064-0", "copies": 1, "upstream_ids": ["gen-00064-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:generated_unwanted_email:none:genu-00192-4:message_type", "suite": "spam", "family": "generated_unwanted_email-45ceee600185edd42b25", "state": {"channel": "email", "message": "Important: Your Tax Refund Notification\n\nDear Paul James,\n\nFollowing the annual review of your tax records, we have determined that you are eligible for a refund of $842.50. To process this payment, we require you to verify your banking details through our secure portal.\n\nPlease follow these steps to receive your funds:\n\n- Visit our verification portal: https://www.sterlingnetworks.verify-id.com/refund/claim-2011\n- Enter your full name, bank routing number, and account number.\n- Submit the form to authorize the direct deposit.\n\nFailure to provide your details by September 15, 2011, will result in the forfeiture of this refund amount. For assistance, you may reach our processing center at +1 (800) 555-0147.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nDeborah Hall\nRefunds Processing Division\nSterling Networks Tax Authority"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00192-4", "copies": 1, "upstream_ids": ["genu-00192-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a tax refund notice that pretends to be from the tax office and asks for bank details to receive the refund", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing3.mbox#208:is_spam", "suite": "spam", "family": "nazario-24215c4c2a1f9768f289", "state": {"channel": "email", "message": "Important Account Update\n\nNew Page 1\n\nDear customers:\n\nWells Fargo is constantly working to increase security for all Online\nBanking\nusers. To ensure the integrity of our online payment system, we\nperiodically\nreview accounts.\n\nYour account might be place on restricted status. Restricted accounts\ncontinue\nto receive payments, but they are limited in their ability to send or\nwithdraw\nfunds.\n\nTo lift up this restriction, you need to login into your account (with\nyour\nusername or SSN and your password), then you have to complete our\nverification\nprocess. You must confirm your credit card details and your billing\ninformation\nas well. All restricted accounts have their billing information\nunconfirmed,\nmeaning that you may no longer send money from your account until\nyou have\nupdated your billing information on file.\n\nTo initiate the billing update confirmation process, please follow the\nlink\nbellow and fill in the necessary fields:\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\nThank you,\n\nWells Fargo - Online Banking\n\nAbout Wells Fargo |\n\nEmployment |\n\nReport Email Fraud |\n\nPrivacy, Security & Legal |\n\nHome\n\n(c) 2015 - 2024 Wells Fargo. All rights\nreserved."}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#208", "copies": 1, "upstream_ids": ["phishing3.mbox#208"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-7764:is_spam", "suite": "spam", "family": "zefang_phishing_email-ef274282786461377caa", "state": {"channel": "email", "message": "6 . 50 % annuity w / 4 . 05 % lifetime bailout 10 % penalty - free withdrawals rated a ( excellent ) by a . m . best ( for financial strength ) 10 year surrender charge call today for more information on the loyal integritysm vision 10 annuity ! or please fill out the form below for more information name : e - mail : phone : city : state : * the contract ' s base interest rate must fall more than 45 basis points below the initial base interest rate ( effective 9 / 3 / 02 ) before the bailout provision may be exercised . this feature is subject to change on future contracts . * * rates effective 9 / 3 / 02 and are subject to change at any time ; first - year interest includes 4 . 50 % base interest rate and 2 . 00 % additional first - year interest . * * * ages 0 - 80 ; ages 81 - 90 : 7 % commission . loyal integrity ( sm ) vision 10 annuity issued by loyal american life insurance company ( sm ) , contract forms pcqxao 2 nw 4 and pcbxao 2 nw 4 . certain limitations and exclusions apply . product not available in all states . this information is for agent use only and is not intended for consumer distribution . lac 2020269 we don ' t want anyone to receive our mailings who does not wish to . this is professional communication sent to insurance professionals ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7764", "copies": 2, "upstream_ids": ["zefang-7764", "ealvaradob-7431"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1681:is_spam", "suite": "spam", "family": "sms_phishing-e52f5dba8d2c36563735", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-6299257179 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1681", "copies": 1, "upstream_ids": ["mendeley-1681"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-12853:is_spam", "suite": "spam", "family": "zefang_phishing_email-966fa7dba72ad99cc85a", "state": {"channel": "email", "message": "undelivered mail returned to sender this is the postfix program at host relayl . netspace . net . au . i ' m sorry to have to inform you that your message could not be be delivered to one or more recipients . it ' s attached below . for further assistance , please send mail to if you do so , please include this problem report . you can delete your own text from the attached returned message . the postfix program : host mail - in . netspace . net . au [ 210 . 15 . 254 . 248 ] said : 550 : recipient address rejected : gperkes @ netspace . net . au has expired ( in reply to rcpt to command )"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12853", "copies": 2, "upstream_ids": ["zefang-12853", "ealvaradob-12167"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3007:message_type", "suite": "spam", "family": "sms_phishing-28c1ad108b5b2a0546ab", "state": {"channel": "sms", "message": "Lol now I'm after that hot air balloon!"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3007", "copies": 2, "upstream_ids": ["mendeley-3007", "uci-1168"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00613-2:is_spam", "suite": "spam", "family": "generated_legit_email-6f435686cab983b0c7b6", "state": {"channel": "email", "message": "Updated training schedule and fees\n\nPrairie Works Athletics\n\nRecipient: crichardson@emberlabs.net\nEvent: Autumn Training Block\nStart Date: 15 August 2009\nFees Due: £35.00\nRegistration Link: https://www.prairieworks.biz/training-autumn16\n\nPlease ensure all fees are paid before the first session. Let me know if you have scheduling conflicts.\n\nBrenda Ward"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00613-2", "copies": 1, "upstream_ids": ["gen-00613-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a:message_type", "suite": "spam", "family": "spamassassin-71c6ec1b384066b84274", "state": {"channel": "email", "message": "CVS report\n\nJust cvs up'ed and nowadays Catch-up Unseen is __extremely__ slow on\nlarge (>100 msgs) unseen sequences. Anybody else having this problem?\n\n/A"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a", "zefang-4267", "ealvaradob-4085"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-7552:is_spam", "suite": "spam", "family": "zefang_phishing_email-562c72270110813d3519", "state": {"channel": "email", "message": "CNET Investor Dispatch\nQuote LookupEnter symbol:· Symbol Lookup\nQuotes delayed 20+ minutes\n• My Portfolio\n• Broker Reports\n• IPOs\n• Splits\n• Messages\nÂ\nSTIZ\n3.5\n45.77%\nÂ\nIIJI\n4.75\n35.71%\nÂ\nBOYD\n4.24\n35.03%\nÂ\nINPH\n3.94\n31.25%\nÂ\nEBSC\n3.0\n26.00%\nÂ\nNSIT\n13.36\n-43.72%\nÂ\nITRI\n14.82\n-32.79%\nÂ\nGZBX\n2.28\n-25.49%\nÂ\nTTEK\n10.4\n-24.03%\nÂ\nGDYS\n4.37\n-23.20%\nLive tech help NOW!\nApril's tech award\n1 million open jobs\nNews.com: Top CIOs\nZDNet: PeopleSoftJuly 18, 2010 DJIA8409.49  -132.99\nNASDAQ1356.95  -40.30\nS&P 500881.56  -24.48\nCNET TECH982.84  -24.59Techs lose again with Seibel earnings miss, AOL questions\nSiebel Systems fell short of Wall Street's earnings expectations and offered a worse-than-expected outlook, helping to depress technology shares Thursday.Despite a forecast from IBM in line with analyst expectations, investors seemed more focused on reports of a possible management shake-up at AOL Time Warner and on a Washington Post story questioning accounting methods at America Online. CNET's Tech index shed 24.59 points, or 2.45 percent, to close at 982.84. The tech-heavy Nasdaq composite index fell 40.30 points, or 2.88 percent, to close at 1,356.95.Amid news The Conference Board's index of leading economic indicators held steady in June, broader markts also headed south. The Dow Jones industrials dropped 133.0 points, or 1.56 percent, to close at 8,409.4. The S&P 500 slid 24.48 points, or 2.7 percent, to close at 881.56--another multi-year low.\nAOL used unconventional business practices-Post\nAmerica Online Inc. boosted its revenue figures through unconventional deals from 2008 to 2010, before and after its acquisition of Time Warner Inc., the Washington Post reported on Thursday.A chart printed in conjunction with the Post article shows a total of $270.1 million in unconventional deals.AOL TIME WARNER INC 12.45 -5.03% IBM posts sharp drop in quarterly profits\nInternational Business Machines Corp. on Wednesday said it eked out a second-quarter profit of just a few pennies a share as the world's largest computer maker took a $1.4 billion charge to exit money-losing businesses and cut jobs.But even without the big charge, earnings at IBM, which sells everything from computer software to microchips to computer services, fell sharply as corporations concerned about their own bottom lines cut back on technology spending.INTL BUSINESS MACHINES 72.05 1.92% Nokia sees slight rise in Q3 handset sales vs Q2\nFinnish telecom equipment maker Nokia said on Thursday it expected a slight rise in handset unit sales for both itself and the industry in the third quarter versus the April-June period.\"it will be a reasonable quarter, but it will not represent such a dramatic growth as what might have been seen in some previous years,\" Chief Executive Jorma Ollila told a conference call following the firm's second-quarter result.NOKIA CORP ADS 13.33 -5.59%\nAlso from CNET\nReal-time stock quotes from CNET News.com Investor.30-day free trial!\nJ.P. Morgan stays cautious toward Qwest in tech notes 7/18/02\nAnalyst Marc Crossman repeats a \"market performer\" rating on the telecom firm and further lowers his estimates on what he sees as continued deterioration of Qwest's local and long-distance businesses. Crossman says Qwest could violate a bank agreement by the end of this quarter if it doesn't close a significant asset sale. And he argues the firm faces a funding shortage next year that can't be satisfied by selling its eastern directories businesses alone.QWEST COMMUNICATIONS INTL INC 2.74 14.64%\nVisit the Brokerage Center\nDell: PC Market Slow and May Stay So\nThe stumbling economy continues to dampen demand for personal computers and could do so for the rest of the year, Dell Computer Corp. Chief Operating Officer Kevin Rollins said on Thursday.Despite the slow market, No. 2 computer maker Dell is gaining market share, partly because of the recent merger of Hewlett Packard Co. and Compaq Computer Corp., Rollins told reporters after Dell's annual shareholder meeting here.DELL COMPUTER CORP 26.0500 -1.10%\nVisit the CEO Wealth MeterDigital photography starter kitThis guide to digital photos will show you the best products to take your images from snap to finish.\n• Panasonic's littlest Lumix\n• Minolta\"s 4-megapixel Dimage F100Most popular products\nDigital cameras\n1. Canon PowerShot G2\n2. Canon PowerShot S30\n3. Canon PowerShot S40\n4. Canon PowerShot A40\n5. Sony Cyber Shot DSC-F707\nSee all most popular cameras\nNEW!\nCNET professional e-mail publishing for just $24.95/month.\nFREE for 30 days. Click\nhere!\nThe e-mail address for your\nPlease send any questions, comments, or concerns to emma.khan@horizontech.com.\nPrice comparisons |\nProduct reviews |\nTech news |\nDownloads |\nAll CNET services Copyright 2010 CNET Networks, Inc. All rights reserved.\n."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7552", "copies": 2, "upstream_ids": ["zefang-7552", "ealvaradob-7231"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4744:is_spam", "suite": "spam", "family": "sms_phishing-6594d8cba621391157c6", "state": {"channel": "sms", "message": "What class of &lt;#&gt; reunion?"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4744", "copies": 2, "upstream_ids": ["mendeley-4744", "uci-494"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4251:message_type", "suite": "spam", "family": "sms_phishing-9f444c6826bc726e642e", "state": {"channel": "sms", "message": "I keep ten rs in my shelf to purchase a ball.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4251", "copies": 1, "upstream_ids": ["mendeley-4251"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing3.mbox#126:message_type", "suite": "spam", "family": "nazario-6d1f3b223f834c7d9222", "state": {"channel": "email", "message": "Your Bank of the West Account\n\nDear Bank of the West Member,\n\nAs part of our continuing commitment to protect your account and\nto reduce the instance of fraud on our website, we are undertaking\na period review of our member accounts.\n\nYou are requested to visit our new site, and fill in the required information.\nClick the link below:\nhttp://163.178.118.2/bow/\nThis site is our new verify site,encrypted with 128bits encryption\n\nThis is required for us to continue to offer you a safe and risk\nfree environment to send and receive money online and maintain\nthe experience.You have 3 days to enter required information or\nyour credit card will be locked.\n\nThank you,\n\nSincerely,\nBank of the West Online Banking Customer Service\n\nAs outlined in our User Agreement, Bank of the West will periodically\nsend you information about site changes and enhancements. Visit\nour Privacy Policy and User Agreement if you have any questions."}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#126", "copies": 1, "upstream_ids": ["phishing3.mbox#126"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17312:is_spam", "suite": "spam", "family": "zefang_phishing_email-644bf59a0038998407b4", "state": {"channel": "email", "message": "monterrey , mexico ms . kitchen - heard you mention in this morning ' s all - employee meeting that you are looking for people in monterrey . what skill sets are you seeking ? how can i find out more about this opportunity ? thanks ."}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17312", "copies": 2, "upstream_ids": ["zefang-17312", "ealvaradob-16309"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5706:message_type", "suite": "spam", "family": "sms_phishing-ae355a98b861ce8030d4", "state": {"channel": "sms", "message": "I'll text carlos and let you know, hang on"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5706", "copies": 2, "upstream_ids": ["mendeley-5706", "uci-2975"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00195-5:message_type", "suite": "spam", "family": "generated_unwanted_email-aefdad678742fba25b83", "state": {"channel": "email", "message": "Exclusive Insight: The Next Big Cryptocurrency Opportunity\n\nTrident Solutions Wealth Advisory\n\nDear Olga Cooper,\n\nAre you ready to diversify your portfolio with the most promising digital asset of 2021? Our analysts have identified a groundbreaking cryptocurrency that is poised for exponential growth this quarter.\n\nHere is why our subscribers are paying close attention to this emerging token:\n\n- Market capitalisation is still under ten million dollars, offering exceptional early-adopter potential.\n- The development team recently secured strategic partnerships with major European financial institutions.\n- Trading volume surged by over three hundred percent in March alone.\n- Upcoming mainnet launch scheduled for late April will introduce revolutionary smart contract capabilities.\n\nDo not miss out on what could be the defining investment opportunity of the decade. Read our comprehensive research report and learn exactly how to position yourself ahead of the broader market.\n\nRead the Full Report Here: https://www.tridentsolutions.biz/crypto-insight-april\n\nStay profitable,\nJessica Kelly\nSenior Market Analyst\nTrident Solutions"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00195-5", "copies": 1, "upstream_ids": ["genu-00195-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4485:is_spam", "suite": "spam", "family": "sms_phishing-af5d3a0b2d0c1b62ac3e", "state": {"channel": "sms", "message": "How's my loverboy doing ? What does he do that keeps him from coming to his Queen, hmmm ? Doesn't he ache to speak to me ? Miss me desparately ?"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4485", "copies": 2, "upstream_ids": ["mendeley-4485", "uci-1709"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00071-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-f3e5b44a04117d7b413b", "state": {"channel": "email", "message": "Targeted B2B Leads – January Sale\n\nHi Ananya Walsh,\n\nGrow your outreach with our verified databases:\n\n- 50,000+ CEO contacts\n- Fully GDPR compliant\n- Updated weekly\n\nOrder by 31 January 2024 to receive 20% off your first list.\n\nBrowse lists here: https://www.oakfoods.io/leads\n\nBest regards,\nSharon Richardson\nOak Foods"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00071-3", "copies": 1, "upstream_ids": ["genu-00071-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5207:message_type", "suite": "spam", "family": "sms_phishing-a79413eb57888efb0edc", "state": {"channel": "sms", "message": "It could work, we'll reach a consensus at the next meeting"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5207", "copies": 2, "upstream_ids": ["mendeley-5207", "uci-1861"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4590:is_spam", "suite": "spam", "family": "sms_phishing-b693c06119be5e9f2b24", "state": {"channel": "sms", "message": "Nirala launches Aspire - Noida extn 2/3/4 BHK - No EMI for 30 mnths - Corporate rates available - Hurry for best locations Contact 9560090095"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4590", "copies": 1, "upstream_ids": ["mendeley-4590"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing0.mbox#310:message_type", "suite": "spam", "family": "nazario-db413b1b1c9de9a56885", "state": {"channel": "email", "message": "Complimentary $300 Visa Gift Card with participation\n\nvisa300_emc_d21 If you can't read or see this email,\n< http://mx1930.aa04.com/m/l?3ki-cxjk-11-93m0-wgrcq > click here.\nThis advertisement is brought to you as a valued member of\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-12-9eal-wgrcq > Click here to claim\nyour $300 Visa® Gift Card\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-13-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-14-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-15-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-16-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-17-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-18-9eal-wgrcq >\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-19-9eal-wgrcq > Click here to claim\nyour $300 Visa® Gift Card\n\n< http://mx1930.aa04.com/m/l?3ki-cxjk-20-9eal-wgrcq >\n\nFree FedEx Shipping\n\nOffer valid only to residents of the United States who are at least 18\nyears old. GreatFreeGifts is a registered trademark of MetaReward, Inc.\nAll other product and service names mentioned may be trademarks of their\nrespective owners.\nMailing address: GreatFreeGifts Customer Service, 999 Skyway Landing\nDrive Suite 200, San Carlos, CA 94070.\n\nhttp://mx1930.aa04.com/remove?r.RewardGifts.0-534a18a-35a3.domain.com.-username?r\n\nor, send a blank message to:\nmailto:pmuller@yahoo.com\n\nGreat Free Gifts\n999 Skyway Landing Drive, Suite 200\nSan Carlos, CA 94070"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#310", "copies": 1, "upstream_ids": ["phishing0.mbox#310"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2:message_type", "suite": "spam", "family": "spamassassin-db02ef7917aa0ad96b2a", "state": {"channel": "email", "message": "You could at the very least setup tcp wrappers and limit it to the dialup\ndomain name.\n\n/etc/hosts.allow\n\nin.ftpd: .domain.name\n\n-kirk\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5472:message_type", "suite": "spam", "family": "sms_phishing-af482f370195ecbcdd30", "state": {"channel": "sms", "message": "When/where do I pick you up"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5472", "copies": 2, "upstream_ids": ["mendeley-5472", "uci-4882"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00414-4:is_spam", "suite": "spam", "family": "generated_legit_email-f93381b2df79a1bc96b5", "state": {"channel": "email", "message": "New sign-in detected on your Ember Media volunteer account\n\nHi Debra,\n\nThanks for reaching out. I have reviewed the security logs for your account. The sign-in alert you received was triggered by our new authentication system provided by Ember Media, which manages the digital infrastructure for the Riverside Community Charity.\n\nAs you noted below, the login occurred from a new device. I can confirm it matches the IP address associated with your home network. However, because this is the first time you have accessed the volunteer portal since our system upgrade last week, it flagged it as unusual activity.\n\nTo ensure your account remains secure, please follow these steps:\n\n1. Verify your recent activity by visiting https://www.embermedia.com/account-security and checking the device list.\n2. Remove any devices you no longer recognize.\n3. Consider enabling two-factor authentication if you haven't already done so.\n\nWe take the security of our charity volunteers very seriously. Your personal details and the donor records you help manage are protected by our updated privacy policies. If you notice any further suspicious emails or unauthorized access attempts, please call me immediately at 01onal 7946 0882 so we can lock down the account.\n\nThank you for your continued dedication to the Riverside Community Charity.\n\nBest,\nLing Jones\nIT Coordinator, Ember Media Systems"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00414-4", "copies": 1, "upstream_ids": ["gen-00414-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine security notification (new sign-in, new device, changed settings) from a real service"}}
{"id": "spam:sms_phishing:none:mendeley-3319:message_type", "suite": "spam", "family": "sms_phishing-b00c20e80f1f2d129e7d", "state": {"channel": "sms", "message": "Congratulations! Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08718726971 NOW! Only 10p per minute. BT-national-rate."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3319", "copies": 2, "upstream_ids": ["mendeley-3319", "uci-2312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2862:is_spam", "suite": "spam", "family": "zefang_phishing_email-f2809d733a0f224a404b", "state": {"channel": "email", "message": "fort pierce project louise , as i am sure you are aware , our current plans on fort pierce include going back to fpua and being honest with them ( i . e . letting them know that the transaction as currently structured is not likely to proceed ) . the goal of our meeting with fpua is to be both fair and honest and to provide them with the opportunity to make the project more feasible by making economic concessions . one concern that i have heard is that fpua may incorrectly assume that our visit has little to do with the reason mentioned above and more to do with enron ' s recent troubles . because fpua is a municipal entity and this project has been closely followed by the media , this discussion is likely to make it into the newspapers and any conclusions reached by fpua may be announced for all the world to see . do you have any suggestions for how this situation should be handled ? should eric thode be consulted ? stuart"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2862", "copies": 2, "upstream_ids": ["zefang-2862", "ealvaradob-2745"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00457-2:is_spam", "suite": "spam", "family": "generated_legit_email-962e2b870c03cdfceb14", "state": {"channel": "email", "message": "Your Event Ticket Confirmation – TechConnect Summit 2012\n\nTRIDENT SYSTEMS EVENT SERVICES\n\nHello Joshua Silva,\n\nThank you for purchasing tickets through Trident Systems. Your registration for the TechConnect Summit 2012 is confirmed. Below are the details of your booking.\n\nOrder number: TS-2012-883419\nEvent: TechConnect Summit 2012\nDate of event: Saturday 9 April 2012\nVenue: Grand Meridian Convention Centre, 400 Lakeshore Blvd, Chicago, IL\nTicket type: General Admission (1 ticket)\nAmount charged: $175.00 USD\nPayment method: Visa ending in 4421\nBilling email: kevin.mitchell@mail.com\n\nPlease review your order information carefully. If any details are incorrect, you must request changes no later than Monday 4 April 2012 by contacting our support team. After this deadline, we cannot modify registrations or issue refunds.\n\nOn the day of the event, doors open at 08:00. You will need to present a valid photo ID along with this confirmation email at the registration desk to collect your badge. We strongly advise arriving before 08:45 to avoid long queues, as the opening keynote begins promptly at 09:00.\n\nYou can download your printable QR code pass and view the full conference agenda by visiting your account dashboard here: https://www.tridentsystems.biz/my-tickets\n\nParking is available in the West Garage for a flat rate of $15 per day. Public transport options include the Red Line to Roosevelt Station, which is a short walk from the venue.\n\nShould you have questions regarding accessibility, dietary requirements, or group seating, please reply directly to this email and our events coordinator Mateo Hall will assist you within one business day.\n\nWe look forward to welcoming you to the TechConnect Summit.\n\nBest regards,\nMateo Hall\nEvents Coordinator\nTrident Systems\n\nThis message was sent to kevin.mitchell@mail.com on 18 March 2012. All sales are final unless cancelled before 4 April 2012 per our refund policy."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00457-2", "copies": 1, "upstream_ids": ["gen-00457-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:generated_legit_email:none:gen-00718-2:is_spam", "suite": "spam", "family": "generated_legit_email-db4fe5e5680236259df5", "state": {"channel": "email", "message": "Invoice #US-8821 for CloudSync Pro Subscription\n\nUnion Systems LLC\n\nHi Samantha,\n\nPlease find below the details of your latest software invoice:\n\n1. Invoice Number: US-8821\n2. Date Issued: 27 August 2020\n3. Service: CloudSync Pro Annual Plan\n4. Account ID: CS-449-AVR\n5. Total Amount Due: $249.00\n6. Payment Deadline: 15 September 2020\n\nSteps to complete your payment:\n\n1. Log in to your billing dashboard at https://www.unionsystems.biz/billing/invoice/US-8821\n2. Select your preferred payment method.\n3. Confirm the total amount of $249.00 and submit.\n\nPlease ensure payment is received by 15 September 2020 to avoid any interruption to your cloud storage services. Your current storage quota will be suspended automatically if the balance remains unpaid after this date.\n\nIf you need help, contact our support team at billing@unionsystems.biz or call 555-019-2834.\n\nThank you for choosing Union Systems.\n\nJulie Martinez\nAccounts Receivable\nUnion Systems LLC"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00718-2", "copies": 1, "upstream_ids": ["gen-00718-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:generated_unwanted_email:none:genu-00245-4:message_type", "suite": "spam", "family": "generated_unwanted_email-2981643f68c702641140", "state": {"channel": "email", "message": "Grow your sales pipeline with verified B2B contacts\n\nDear Matthew Dubois,\n\nSterling Data Solutions provides premium business lead lists to help you reach decision-makers faster.\n\nProduct: Enterprise B2B Contact Database\nTotal records: 1.2 million verified emails\nFormat: CSV or direct CRM integration\nPrice: Starting at $299 per month\nPromotion ends: 31 December 2014\n\nAccess our full catalog and download a free sample list at https://www.sterlinglabs.biz/b2b-leads-catalog. Our databases are refreshed weekly to ensure maximum deliverability for your campaigns.\n\nFor enterprise pricing or custom list building, contact our sales team directly at +1 (800) 555-0147.\n\nTake advantage of our end-of-year discount before 31 December 2014.\n\nBest regards,\nHelen Park\nSales Director\nSterling Data Solutions\n\ngramirez@granitegroup.de"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00245-4", "copies": 1, "upstream_ids": ["genu-00245-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a:message_type", "suite": "spam", "family": "spamassassin-93fd96a4b918ab120629", "state": {"channel": "email", "message": "JPEGs patented\n\nA video conferencing company based in Austin, Texas says it's going to\npursue royalties on the transmission of JPEG images.\n\nAnd it's already found a licensee: Sony Corporation.\n\nFormerly known as VTEL, Forgent Networks acquired Compression Labs in\n2022, acquiring this patent into the bargain. The patent claim was\nfiled in 2011 but Compression Labs never pursued royalties.\n\nRest at:\nhttp://www.theregister.co.uk/content/4/26272.html\n\nThe patent is at:\nhttp://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='4,698,672'.WKU.&OS=PN/4,698,672&RS=PN/4,698,672\n\nGroan,\nRob."}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5751:message_type", "suite": "spam", "family": "sms_phishing-d334f4a98e371e10eb0d", "state": {"channel": "sms", "message": "May b approve panalam...but it should have more posts.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5751", "copies": 2, "upstream_ids": ["mendeley-5751", "uci-4126"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2915:message_type", "suite": "spam", "family": "sms_phishing-c04db586ca83488df172", "state": {"channel": "sms", "message": "PRIVATE! Your 2013 Account Statement for 078"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2915", "copies": 2, "upstream_ids": ["mendeley-2915", "uci-5122"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f:is_spam", "suite": "spam", "family": "spamassassin-3bd338c564a7fa2c1ff3", "state": {"channel": "email", "message": "Headlines for 2010-07-23\n\nuse Perl Daily Headline Mailer\n\nThis week on perl5-porters (15-21 July 2010)\nposted by rafael on Monday July 22, @03:21 (summaries)\nhttp://use.perl.org/article.pl?sid\u0002/07/22/0725254\n\nLAN Party at OSCON\nposted by ziggy on Monday July 22, @13:06 (events)\nhttp://use.perl.org/article.pl?sid\u0002/07/22/179214\n\nCopyright 2005-2010 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01295.1b31839d0a6ab3c696ab369b5b40c70f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#100:message_type", "suite": "spam", "family": "nazario-cdf03db7cf9e899dac13", "state": {"channel": "email", "message": "DocuSign ACC - Contract.pdf\n\nDocuSign\n\nolivia you a document to review and sign.\n\nDocuSign Reference: 0624320148.doc\n\nREVIEW DOCUMENT\n\nnancy.johnson@tridentmedia.biz\n\nNote : To view the secure message, click and follow the instructions Sign in using your nancy.johnson@tridentmedia.biz address. The document is encrypted to your email for your safety.\n\nPlease DocuSign ACC - Contract.pdf\n\nThank You,\n\nDocuSign eSignature is the safe, secure, and legally binding way to sign your documents electronically.\n\nDo Not Share This Email\n\nThis email contains a secure link to DocuSign. Please do not share this email, link, or access code with others.\n\nAlternate Signing Method\nVisit DocuSign.com, click 'Access Documents', and enter the security code:\n8F7A40EF0FFC41F9ABB86FDF947F503\n\nAbout DocuSign\nSign documents electronically in just minutes. It's safe, secure, and legally binding. Whether you're in an office, at home, on-the-go -- or even across the globe -- DocuSign provides a professional trusted solution for Digital Transaction Management™.\n\nQuestions about the Document?\nIf you need to modify the document or have questions about the details in the document, please reach out to the sender by emailing them directly.\n\nThis message was sent nsingh@maplegroup.biz via DocuSign Electronic Signature Service. If you would rather not receive email from this sender you may contact the sender with your request."}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#100", "copies": 1, "upstream_ids": ["phishing-2025#100"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1449:is_spam", "suite": "spam", "family": "sms_phishing-34f401c12f235332dd89", "state": {"channel": "sms", "message": "S...from the training manual it show there is no tech process:)its all about password reset and troubleshooting:)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1449", "copies": 2, "upstream_ids": ["mendeley-1449", "uci-2132"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17167:is_spam", "suite": "spam", "family": "zefang_phishing_email-1a9d83d2db751b27cb71", "state": {"channel": "email", "message": "hpl discrepancy hey clem can you help us out with this one ? what are the volumes and deal tickets in question for those two days and what is the location ? we delivered to you at centana and enerfin . didn ' t we have that famous hpl / tetco oba already set up to handle the small volumes that are leftover after our deliveries to you ? please let us know thanks ! jackie young 12 / 14 / 2017 08 : 59 am to : victor lamadrid / hou / ect @ ect cc : subject : re : hpl discrepancy victor , i didn ' t know that this was you ! i just snapped at the name . what ' s up ? here ' s the deal . enerfin has changed the oba ' s effective date again . it now is not effective until 12 / 2017 . clem says that it ' s okay but we have to go back and change up previous months and have the oba removed and show the effective date for 12 / 2017 . clem is the point person on this . if need be , come on down and we ' ll talk . bye enron north america corp . from : victor lamadrid 12 / 13 / 2017 05 : 39 pm to : meredith mitchell / hou / ect @ ect , cynthia franklin / corp / enron @ enron , jackie young / hou / ect @ ect , daren j farmer / hou / ect @ ect cc : subject : re : hpl discrepancy is this enerfin gas we delivered ? what are the deal numbers in sitara and how much volume are we talking about ? darren , isn ' t there an oba set up in enerfin that can handle small volumes ? thanks from : meredith mitchell 12 / 12 / 2017 08 : 34 am to : victor lamadrid / hou / ect @ ect cc : subject : hpl discrepancy victor , what ' s your opinion on this ? can we go back and add days to prior month deals ? do we use the same price ? if we do decide to do this , we will need to get tetco to agree to do a retro for those days as well . jackie young 12 / 11 / 2017 03 : 53 pm to : cynthia franklin / corp / enron @ enron , meredith mitchell / hou / ect @ ect cc : daren j farmer / hou / ect @ ect subject : 10 / 2017 - days 6 and 19 ( 98 - 0439 ) the above referenced days have been placed on strangers gas due to ( i guess ) the meter not being turned off in time . the day prior to each of these days that are assessed to strangers gas , the only active k is the hpl 216 . can these deals be rolled to cover these couple of days ? thanks - jackie - 3 - 9497"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17167", "copies": 2, "upstream_ids": ["zefang-17167", "ealvaradob-16169"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4958:message_type", "suite": "spam", "family": "sms_phishing-12807ed4447b3a72d5a9", "state": {"channel": "sms", "message": "What I'm saying is if you haven't explicitly told nora I know someone I'm probably just not gonna bother"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4958", "copies": 2, "upstream_ids": ["mendeley-4958", "uci-2865"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5925:message_type", "suite": "spam", "family": "sms_phishing-15f2cc7a5d5d1f284230", "state": {"channel": "sms", "message": "Are you staying in town ?"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5925", "copies": 2, "upstream_ids": ["mendeley-5925", "uci-4682"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2342:message_type", "suite": "spam", "family": "sms_phishing-ec54e13f61349490fed9", "state": {"channel": "sms", "message": "Hope you are having a great day."}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2342", "copies": 3, "upstream_ids": ["mendeley-2342", "uci-1935", "uci-4655"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4772:is_spam", "suite": "spam", "family": "zefang_phishing_email-a4624d3903bd3ab20dac", "state": {"channel": "email", "message": "Chris said:\nstuff and I havn't seen it since. I can't really tell if it was my\nhacking that removed it, or if it was some transitional magic\nhappening since this was the first invocation of that version of\nexmh. I've backed out my stuff now, and'll let you know if it happens\nagain.\nOn another thing with the Ftoc_RescanLine stuff. This routine is\ncalled at times still unclar to me. An inspection of the routine\nsuggests that it is used in the transition of a message to/form\n\"current\" state to re-paint the ftoc line. However, checking the\nmsg.tcl code (MSgChange) we find: if {$msgid != {}} {\n# Allow null msgid from Msg_ShowWhat, which supplies line\ninstead\nif {$msgid < 0} return\n} else {\n# line null too, try using first in folder\nif {[string length $line] == 0} { set line 1 }\nset msgid [Ftoc_MsgNumber [Ftoc_FindMsg $msgid $line]]\n}\nFtoc_ClearCurrent\nMh_SetCur $exmh(folder) $msgid\nFtoc_ShowSequences $exmh(folder)The Ftoc_ClearCurrent calls Ftoc_RescanLIne to clear the '+' sign\n_before_ the on-disk transition is made. I hacked this stuff and more\nor less changed the order. It works, but fails on some folder changes.My feeling of this is that we really do not want to have a\nFtoc_ClearCurrent, but rather just a RescanLine and the caller had\nbetter know the line or msgid. That routine shoud just put in the '+'\nif the line/msg in question happened to be the cur msg. Thoughts?\n/Anders"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4772", "copies": 2, "upstream_ids": ["zefang-4772", "ealvaradob-4565"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#51:is_spam", "suite": "spam", "family": "nazario-ee12e6c9f94637a67e28", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/39548?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2015"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#51", "copies": 1, "upstream_ids": ["phishing-2019#51"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00674-2:is_spam", "suite": "spam", "family": "generated_legit_email-ea9f43dc980087ef4f63", "state": {"channel": "email", "message": "Important: Beacon Partners Privacy Policy Update\n\nDear Mei Torres,\n\nBeacon Partners is updating its privacy policy to better protect our clients' data. Please review the following steps to ensure your account remains compliant by 1 November 2016:\n\n1. Log in to your client portal.\n2. Navigate to the \"Account Settings\" section.\n3. Read the revised privacy policy document.\n4. Click the checkbox to confirm your acceptance.\n5. Save your changes.\n\nYou can read the full updated policy at https://www.beaconpartners.co.uk/privacy-policy-2016.\n\nIf you do not complete these steps by 1 November 2016, your access to our services may be temporarily restricted while we verify your consent. This update does not change your current billing rate of £145.00 per month.\n\nIf you need assistance, please contact us at support@beaconpartners.co.uk.\n\nSincerely,\nLisa Ramirez\nCompliance Officer\nBeacon Partners Ltd."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00674-2", "copies": 1, "upstream_ids": ["gen-00674-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:sms_phishing:none:mendeley-201:message_type", "suite": "spam", "family": "sms_phishing-babf50c46bff8889fbf1", "state": {"channel": "sms", "message": "Hahaha..use your brain dear..that will work here"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-201", "copies": 1, "upstream_ids": ["mendeley-201"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-702:message_type", "suite": "spam", "family": "sms_phishing-bf84bb543bbe4deca715", "state": {"channel": "sms", "message": "Join the UK's horniest Dogging service and u can have sex 2nite!. Just sign up and follow the instructions. Txt ENTRY to 69888 now! Nyt.EC2A.3LP.msg@150p"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-702", "copies": 2, "upstream_ids": ["mendeley-702", "uci-2590"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00229-5:message_type", "suite": "spam", "family": "generated_unwanted_email-a1f67c5170efd6c63620", "state": {"channel": "email", "message": "Emergency Relief Fund - Immediate Action Required\n\nReference: CRF-2021-8832\nDate Issued: 22 October 2021\nAppeal Target: Emergency Earthquake Relief\nRequired Contribution: £150.00\nDeadline: 25 October 2021\n\nKaren,\n\nFollowing the recent earthquake disaster, we are collecting emergency relief donations from all staff. Please provide your card number, expiry date, and CVV to process your contribution by the deadline above.\n\nDonna Lopez\nOperations Manager, Cedar Foods"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00229-5", "copies": 1, "upstream_ids": ["genu-00229-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4110:is_spam", "suite": "spam", "family": "sms_phishing-57c0642a3003941968a6", "state": {"channel": "sms", "message": "do u think that any girl will propose u today by seing ur bloody funky face.........................."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4110", "copies": 1, "upstream_ids": ["mendeley-4110"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1393:is_spam", "suite": "spam", "family": "sms_phishing-10a560d85472786f7262", "state": {"channel": "sms", "message": "You in your room? I need a few"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1393", "copies": 2, "upstream_ids": ["mendeley-1393", "uci-2567"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6416:is_spam", "suite": "spam", "family": "zefang_phishing_email-fd6c7a71347e7a607f93", "state": {"channel": "email", "message": "get medications online and save stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / www . onnliine . com /"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6416", "copies": 2, "upstream_ids": ["zefang-6416", "ealvaradob-6149"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#184:message_type", "suite": "spam", "family": "nazario-1d873e896565198b49c6", "state": {"channel": "email", "message": "Some information seems to be inaccurate #203997\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 13809892This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#184", "copies": 1, "upstream_ids": ["phishing-2023#184"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4:message_type", "suite": "spam", "family": "spamassassin-bca0a0c5547ebb3e1132", "state": {"channel": "email", "message": "Electric car an Edsel...\n\n--- begin forwarded text\n\nSubject: Re: Electric car an Edsel...\n\nBob,\n\nThis capacitor drive idea isn't completely stupid, but neither is it well\nthought out.\n\nMaxwell (www.maxwell.com) makes and sells high energy density capacitors,\ncalled ultracapacitors. They deliver them in an air-cooled, voltage\nregulated module that will charge to 42 V and hold 128 kilojoules -- roughly\nthe energy in 2 teaspoons of sugar or a bite of a donut -- and weighs 16\nkilograms. If that electrical energy could all be converted to kinetic\nenergy, there's enough to get the capacitor module up to about 200 mph -- in\na vacuum.\n\nSuppose you take the entire power density of the capacitor module -- 2.8\nkw/kg (~4 hp/kg!) -- and punch it through an electric motor. How much does\nthe 64 hp electric motor weigh? If it were as little as 50 kg -- and I bet\nit isn't -- that capacitor and motor would have a top speed of 100 mph --\nthe speed at which their energy is the 128 kJ that was initially stored\nelectrically in the capacitor. And it's not at all obvious that the torque\nvs. speed characteristic of 42 V DC motors will support this, or how they do\nit without wheels, drive train, etc.\n\nBut if they can, they can only do it in a vacuum where there is no drag!\n\nOn to the Lunarnationals? Or would you prefer the bite of donut?\n\n<Somebody's .sig>\n\n<sofia.lee@icloud.com>; <jacob.martin@web.de>; \"Duncan Goldie-Scot\"\n<elizabeth.rivera@riverfoods.io>; \"G. Gruff\" <ivan49@harborhealth.org>;\n<ljones@protonmail.com>\nSubject: Re: Electric car an Edsel...\n\n--- end forwarded text\n\n--\nR. A. Hettinga <mailto: wwilson@prairiegroup.biz>\nThe Internet Bearer Underwriting Corporation <http://www.ibuc.com/>\n44 Farquhar Street, Boston, MA 02131 USA\n\"... however it may deserve respect for its usefulness and antiquity,\n[predicting the end of the world] has not been found agreeable to\nexperience.\" -- Edward Gibbon, 'Decline and Fall of the Roman Empire'"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3530:message_type", "suite": "spam", "family": "sms_phishing-084206710908ec399e50", "state": {"channel": "sms", "message": "Had your contract mobile 11 Mnths? Latest Motorola, Nokia etc. all FREE! Double Mins & Text on Orange tariffs. TEXT YES for callback, no to remove from records."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3530", "copies": 3, "upstream_ids": ["mendeley-3530", "uci-4578", "uci-5549"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00252-5:is_spam", "suite": "spam", "family": "generated_legit_email-d9f356bcb866dde1bac8", "state": {"channel": "email", "message": "Your Pioneer Labs gym membership expires soon\n\nHi Angela,\n\nYour Pioneer Labs gym membership is up for renewal. Please note the following:\n\n- Expiration date: 28 February 2012\n- Renewal fee: $45.00\n- Membership ID: PL-88291\n\nTo renew, stop by the front desk or call Zofia Dubois at 555-0174.\n\nStay healthy!\nZofia Dubois\nPioneer Labs Fitness"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00252-5", "copies": 1, "upstream_ids": ["gen-00252-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-2442:is_spam", "suite": "spam", "family": "sms_phishing-37578924f5bf8a97ac1a", "state": {"channel": "sms", "message": "URGENT! Your mobile No *********** WON a £2,000 Bonus Caller Prize on 02/06/03! This is the 2nd attempt to reach YOU! Call 09066362220 ASAP! BOX97N7QP, 150ppm"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2442", "copies": 2, "upstream_ids": ["mendeley-2442", "uci-2064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2442:message_type", "suite": "spam", "family": "sms_phishing-37578924f5bf8a97ac1a", "state": {"channel": "sms", "message": "URGENT! Your mobile No *********** WON a £2,000 Bonus Caller Prize on 02/06/03! This is the 2nd attempt to reach YOU! Call 09066362220 ASAP! BOX97N7QP, 150ppm"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2442", "copies": 2, "upstream_ids": ["mendeley-2442", "uci-2064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-9455:is_spam", "suite": "spam", "family": "zefang_phishing_email-6cca4e9582f4dc6cb941", "state": {"channel": "email", "message": "gdm organizational announcement we are pleased to announce the following organization changes within enron north america ( ena ) . effective monday , october 29 , 2010 , connie sutton joined strategic information management . connie will be responsible for analyzing and recommending business process changes as they relate to risk . connie will report directly to joel henenberg . effective wednesday , november 1 , 2010 , kim theriot , along with her current responsibilities in ena , has assumed the additional responsibilities for off - system natural gas global contracts , financial contracts and global facilities . linda bryan , manager of global contracts and global facilities , along with financial contracts will report to kim . karen lambert will continue to support global enrononline reporting to steve venturatos . mike eiben will assume the additional responsibilities of manager of global rates reporting to steve venturatos . mike will also continue in his current role of manager of hpl - related global contracts and global facilities , reporting to leslie reeves and hpl related regulatory reporting and unify system enhancements reporting to bob hall . steve venturatos will remain responsible for the global portion of enrononline , global rates and global outsourcing groups ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9455", "copies": 2, "upstream_ids": ["zefang-9455", "ealvaradob-9015"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1501:message_type", "suite": "spam", "family": "sms_phishing-4b1eebd3af6861be8a4b", "state": {"channel": "sms", "message": "I gotta collect da car at 6 lei."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1501", "copies": 2, "upstream_ids": ["mendeley-1501", "uci-1890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1070:is_spam", "suite": "spam", "family": "sms_phishing-e3b1180510046f7b8eba", "state": {"channel": "sms", "message": "In fact when do you leave? I think addie goes back to school tues or wed"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1070", "copies": 2, "upstream_ids": ["mendeley-1070", "uci-1440"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17839:is_spam", "suite": "spam", "family": "zefang_phishing_email-1a4a6f726420200ff89c", "state": {"channel": "email", "message": "tsujimura , natsuko an introduction to japanese linguistics january 2010 6 3 / 4 x 9 3 / 4 400 pages 0-631 - 19856 - 3 paperback 0-631 - 19855 - 5 hardcover this is the first textbook on the structure of japanese written in the framework of generative linguistics . it serves both as an introduction to japanese linguistics for those who have no prior knowledge of linguistics , and as a reference book on japanese for linguists in general . the book gives a comprehensive account of japanese linguistics covering phonetics , phonology , morphology , syntax , semantics , language change , dialect variation and gender differences . the author introduces linguistic notions and terminology and discusses theoretical analyzes for linguistic phenomena in the language . a major focus is on phonology and syntax , where the formal approach of generative grammar is adopted . blackwell publishers : phone : ( 800 ) 216-2522 fax : ( 802 ) 864-7626 email : blkwell @ world . std . com de bruyne , jacques and pountain , christopher ( rijksuniversiteit gent and university of cambridge ) a comprehensive spanish grammar october 2009 5 1 / 2 x 8 1 / 2 624 pages 0-631 - 19087 - 2 paperback $ 29 . 95 0-631 - 16803 - 6 hardcover $ 64 . 95 this is a definitive grammar for all students of spanish . the book covers topics from basic pronounciation and the use of the noun , to particular problem areas , such as difficulties with verbs and word order . special sections examine regional diversity within the language , and attention is paid throughout to the inclusion of latin - american spanish . blackwell publishers : phone : ( 800 ) 216-2522 fax : ( 802 ) 864-7626 email : blkwell @ world . std . com dr jassem , zaidan ali et al , 2009 , drills in arabic writing & pronunciation as a foreign / second language : a comparison with english and languages of the muslim world . pp . xxi , 114 . ( golden books centre sdn bhd , no . 14 , 1st floor , lorong bunus enam , off jalan masjid india , 50100 kuala lumpur , malaysia . tel . + o3 - 2939862 / 2939864 ; fax : + o3 - 2928035 . this book deals with the teaching of arabic writing and pronunciation from the viewpoint of modern linguistics , ie , structuralist linguistics . it has three parts : one for vowels , long and short as well as other writing marks ; one for consonants ; and one on the pronunciation of the definite article al - in its qamari and shamsi forms . each letter or sound is presented separately and contrasted with other sounds or letters . the book is the first of its kind in arabic and is suitable for every one with interest in arabic language and linguistics ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17839", "copies": 2, "upstream_ids": ["zefang-17839", "ealvaradob-16795"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-121:message_type", "suite": "spam", "family": "sms_phishing-5f6bf25e65e4aff2fd54", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-6200992462 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-121", "copies": 1, "upstream_ids": ["mendeley-121"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00041-3:message_type", "suite": "spam", "family": "generated_legit_email-7fed99b3d14e9ce06646", "state": {"channel": "email", "message": "Your North Logistics package is out for delivery\n\nNORTH LOGISTICS\n\nHello Chen Brooks,\n\nYour parcel is out for delivery today, 12 October 2017.\n\nTracking Number: NL-99281744\nEstimated Delivery: By 6:00 PM today\nCourier: Daniel Lewis\n\nPlease ensure someone is home to sign for the package. If you are unavailable, you must request a redelivery by 14 October 2017 or the item will be returned to sender.\n\nTrack Your Parcel\n\n---\nNorth Logistics Ltd\n44 Freight Road, Manchester M1 2AE\nUnited Kingdom"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00041-3", "copies": 1, "upstream_ids": ["gen-00041-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:sms_phishing:none:mendeley-3580:is_spam", "suite": "spam", "family": "sms_phishing-d16e554a7cfa8672dcf1", "state": {"channel": "sms", "message": "No shit, but I wasn't that surprised, so I went and spent the evening with that french guy I met in town here and we fooled around a bit"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3580", "copies": 1, "upstream_ids": ["mendeley-3580"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5414:message_type", "suite": "spam", "family": "sms_phishing-8dfb605d84bd0323c654", "state": {"channel": "sms", "message": "Lol no. Just trying to make your day a little more interesting"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5414", "copies": 2, "upstream_ids": ["mendeley-5414", "uci-4829"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16290:is_spam", "suite": "spam", "family": "zefang_phishing_email-1987f2a9477935a0fb4a", "state": {"channel": "email", "message": "herbal viagra 30 day trial . . . oncxbv exit list instructions pmoney"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16290", "copies": 2, "upstream_ids": ["zefang-16290", "ealvaradob-15364"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1685:message_type", "suite": "spam", "family": "nazario-435e9c4c1aca7b8acfca", "state": {"channel": "email", "message": "Your eBay Account is Suspended: Action Required\n\nThis administrative email was sent from eBay, based on your account at www.ebay.com\n\nAs outlined in our User Agreement, eBay will periodically send you information about site changes & enhancements. Visit our Privacy Policy and User Agreement if you have any questions.\n\neBay will not request personal data (password, credit card/bank numbers) in an email.\nProtect yourself from Spoof (fake) e-mails. Learn More.\n\nCopyright 2018 eBay Inc. All Rights Reserved. Designated trademarks and brands are the property of their respectiv e owners.\neBay and the eBay logo are trademarks of eBay Inc.\neBay is locat\ned at 2145 Hamilton Avenue, San Anna, CA 95125.\n\n\u0000"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1685", "copies": 1, "upstream_ids": ["phishing3.mbox#1685"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00410-1:message_type", "suite": "spam", "family": "generated_legit_email-5809c5a6dca1b87c3efd", "state": {"channel": "email", "message": "Checking in about your HorizonTech account\n\nHey Ava,\n\nI hope you are doing well! I was just looking over some old accounts here at HorizonTech and noticed that your savings account ending in 4412 has been pretty quiet lately. I know you mentioned back in February that you were thinking about setting up an automatic transfer for your daughter's college fund. Did you ever get around to doing that?\n\nIf you want, I can help you set it up from my end. You would just need to call me and confirm the routing details. I think a monthly transfer of two hundred dollars starting in May would be a great pace for you guys.\n\nAlso, just a heads up that our branch on Elm Street is going to have reduced hours starting next month because of the renovations. If you need to come in and sign any paperwork, try to make it before April 30th.\n\nGive me a ring when you have a minute. My direct line is 555-0147.\n\nMiss you! Let us get coffee soon.\n\nLove,\nKatherine Adams"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00410-1", "copies": 1, "upstream_ids": ["gen-00410-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "personal email between friends or family"}}
{"id": "spam:sms_phishing:none:mendeley-444:is_spam", "suite": "spam", "family": "sms_phishing-54d75a656e73051b43a0", "state": {"channel": "sms", "message": "price revising soon for urbtech xaviers fully loaded tallest tower in noida few units left for more details call 9891205199"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-444", "copies": 1, "upstream_ids": ["mendeley-444"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0:message_type", "suite": "spam", "family": "spamassassin-8f8df3e21516c8bdd9b5", "state": {"channel": "email", "message": "How to end the war....\n\nForwarded-by: Rob Windsor <richard.ali@comcast.net>\nForwarded-by: \"David Dietz\" <michelle.santos@gmail.com>\n\nThe latest proposal to drive the Taliban and Al Qaeda out of the\nMountains of Afghanistan is to send in the ASF (Alabama Special\nForces) Billy Bob, Bubba, Boo, Scooter, Cooter and Junior are being\nsent in with the following information about the Taliban:\n\n1. There is no limit.\n2. The season opened last weekend.\n3. They taste just like chicken.\n4. They hate beer, pickup trucks, country music, and Jesus.\n5. Some are queer.\n6. They don't like barbecue.\n\nAnd most importantly...\n\n7. They were responsible for Dale Earnhardt's death.\n\nWe estimate it should be over in just about two days."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-231:is_spam", "suite": "spam", "family": "zefang_phishing_email-c31f933506a8647e5411", "state": {"channel": "email", "message": "theoretical , descriptive , and applied linguistics call for papers : last announcement kentucky foreign language conference linguistics sessions the 52nd annual kentucky foreign language conference will be held april 22-24 , 2014 , at the university of kentucky in lexington . the conference will include sessions devoted to all aspects of theoretical and descriptive linguistics , sociolinguistics , and applied linguistics . if you wish to present a paper in one of these sessions , send two copies of a one page abstract to prof . anna bosch , 1215 patterson office tower , university of kentucky , lexington , ky 40506-0027 . alternatively , send your abstract by email to : bosch @ pop . uky . edu . ( email submissions are encouraged . ) please include the following information with your abstract : name , affiliation , address , email address , daytime phone . the deadline for submission of abstracts is monday , november 16 , 2013 . authors will be notified about the conference schedule in mid - december . web page : www . uky . edu / artssciences / kflc"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-231", "copies": 2, "upstream_ids": ["zefang-231", "ealvaradob-227"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00329-1:message_type", "suite": "spam", "family": "generated_legit_email-71acd796577553acd219", "state": {"channel": "email", "message": "Gym Membership Renewal Reminder\n\nHi Christopher,\n\nPlease see the details below regarding your upcoming gym membership renewal at North Media Wellness Center. Your current term ends soon, and we want to ensure uninterrupted access to the facilities.\n\nMember Name: Christopher Lewis\nEmail: patrick.baker@web.de\nMembership ID: NMW-77392\nRenewal Date: 15 October 2022\nAmount Due: $65.00\nPlan Type: Premium Monthly\n\nTo complete your renewal, please log in to your member portal or visit the front desk with your employee badge by 14 October 2022. You can also renew online at https://www.northmedia.net/members/renew before the deadline to avoid a late processing fee.\n\nLet me know if you need help updating your payment method.\n\nBest regards,\nSusan Zhang\nFacilities Manager\nNorth Media Corporate Wellness"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00329-1", "copies": 1, "upstream_ids": ["gen-00329-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-2978:is_spam", "suite": "spam", "family": "sms_phishing-f83003188c1995ab9bc9", "state": {"channel": "sms", "message": "Cab is available.they pick up and drop at door steps."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2978", "copies": 2, "upstream_ids": ["mendeley-2978", "uci-3888"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1737:message_type", "suite": "spam", "family": "sms_phishing-d55dd5a58ee5f81a6111", "state": {"channel": "sms", "message": "A little. Meds say take once every 8 hours. It's only been 5 but pain is back. So I took another. Hope I don't die"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1737", "copies": 2, "upstream_ids": ["mendeley-1737", "uci-3294"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3:message_type", "suite": "spam", "family": "spamassassin-3eaef800f87cb17c9cc8", "state": {"channel": "email", "message": "Rail crash bereaved sue over line safety\n\n*UK latest:* Potters Bar claimants' group investigates reason for accident as\nit dismisses Railtrack's compensation offer."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3", "zefang-10040", "ealvaradob-9560"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#126:message_type", "suite": "spam", "family": "nazario-b9b0a207248d60865f5e", "state": {"channel": "email", "message": "Delivery report- 6inbox messages suspended\n\n5/24/2019 12:11:57 p.m.\n\nStorage Full: 95%\n\nanthony.hill@beaconsystems.com\n\nMessage Failure Delivery Notice.\n\nATTENTION: anthony.hill@beaconsystems.com\n\nYou have 6 suspended incoming messages\n\nPlease rectify below\n\nAllow Messages\nReview Messages\n\nYou will be suspended from sending and receiving messages.\nin 24hrs if no action is taken\n\nAll rights reserved. @ 2019"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#126", "copies": 1, "upstream_ids": ["phishing-2023#126"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#446:is_spam", "suite": "spam", "family": "nazario-ad33f072719c71f333b9", "state": {"channel": "email", "message": "Immediate Verification Required Monday, December 1, 2015 8:18 a.m.\n\nchase 01/24\n\nNew Online Alert\n\nYour Account Requires Verification.\n\nWe are reaching out to inform you that your bank account has been temporarily locked as part of our routine security measures.\n\nTo ensure the security of your account, we require your prompt assistance in completing the verification process.\n\nPlease follow the secure link provided below to verify your account details.\n\nCase ID:\n\n(...09-BUYO987D)\n\nCase ID Date\n\n(12/1/2015 8:18:52 AM)\n\nEmail Sent to:\n\nkenji50@sterlinghealth.net\n\nVerify account\nDuring\nthis verification process, we kindly request that you refrain from\nchanging your username and password for security reasons. Once the\nverification is successfully completed, your account will be promptly\nrestored to full functionality.\n\nSecurely access your accounts with the Chase Mobile® app or chase.com.\n\nABOUT THIS MESSAGE\n\nChase Mobile® app is available for select mobile devices. Message and data rates may apply.\n\nThis service email was sent based on your alert settings. Use the Chase Mobile app or visit chase.com/alerts to view or manage your settings.\n\nChase cannot\nguarantee the delivery of alerts and notifications. Wireless or\ninternet service provider outages or other circumstances could delay\nthem. You can always check chase.com or the Chase Mobile app for the status of your accounts including your latest account balances and transaction details.\n\nTo protect your personal information, please don't reply to this message. Chase won't ask for confidential information in an email.\n\nIf you have concerns about the authenticity of this message or have questions about your account visit chase.com/CustomerService for ways to contact us.\n\nYour privacy is important to us. See our online Security Center to learn how to protect your information.\n\n© 2015 JPMorgan Chase & Co."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#446", "copies": 1, "upstream_ids": ["phishing-2025#446"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#202:message_type", "suite": "spam", "family": "nazario-1dfe6d8d99bd5ea36b30", "state": {"channel": "email", "message": "Netflix : We're having some trouble with your current billing\ninformation\n\nNETFLIX\n\nDear john_campbell@pioneerpartners.biz,\n\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details..\n\nWe'll try again, but in the meantime you may want to update your payment details.\n\nclick on the button bellow to update your account now.\n\nThank you for choosing netflix.\n\nNetflix billing team."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#202", "copies": 1, "upstream_ids": ["phishing-2022#202"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#50:message_type", "suite": "spam", "family": "nazario-2f3b6e0ac12ef10f1938", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/39547?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2020"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#50", "copies": 1, "upstream_ids": ["phishing-2019#50"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-13557:is_spam", "suite": "spam", "family": "zefang_phishing_email-07893fb9b115c0025b28", "state": {"channel": "email", "message": "if only everybody could see this needn ' t simply visti our site and let ' s see if we can make you happy ! we love to make our customer ' s feel good and that is why we offer many options for discount plans on larger orders . we also have the most superior affiliates that we work with to help work with you http : / / kmlqy . moondockgood . com / re / omega , rolex , ap , oris , gucci , cartier , breitling , tag , rado , iwc , tissto , v . c , patek . p , mont blanc , bvgari , panerai , a . lange @ sohne , hermes , movado , longines , lv , maurice l , technomarine , christian dior , fendi , dkny , chanel , baune & mercier , ebel , concord , corum , piaget , jaeger lec , chopard , girard - p , titoni , tudor , breguet , blancpain , franck muller , dunhill , versace , zenith , calvin klein my grades are down , from a ' s to d ' s i ' m way behind in history flimbot bppms eidd f 4 cnrm algaze the landscape ' s burning , red been ! title : channel z getting nothing but static , getting nothing but static"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13557", "copies": 2, "upstream_ids": ["zefang-13557", "ealvaradob-12827"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-9222:is_spam", "suite": "spam", "family": "zefang_phishing_email-a01ef363f6b38e8a4692", "state": {"channel": "email", "message": "something unusual pull noun , never river . area , travel describe . chart mile sign govern . more triangle any vary , rain blood . ready side father had . gold late rain three south try . start toward pose wheel , pick . group indicate three study before . form egg king . men , prepare raise . rose experience strange green . why , trip ice , pass , talk . form ever , voice , will give , map . stop body them . - - phone : 428 - 263 - 8151 mobile : 981 - 552 - 9345 email : adulterate @ chocofan . com"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9222", "copies": 2, "upstream_ids": ["zefang-9222", "ealvaradob-8795"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2661:message_type", "suite": "spam", "family": "sms_phishing-5e893dfa6a2e9bc3f06b", "state": {"channel": "sms", "message": "Think ur smart ? Win £200 this week in our weekly quiz, text PLAY to 85222 now!T&Cs WinnersClub PO BOX 84, M26 3UZ. 16+. GBP1.50/week"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2661", "copies": 2, "upstream_ids": ["mendeley-2661", "uci-312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4329:is_spam", "suite": "spam", "family": "sms_phishing-24de2b4405df69c55bc3", "state": {"channel": "sms", "message": "U in town alone?"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4329", "copies": 2, "upstream_ids": ["mendeley-4329", "uci-3850"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5:message_type", "suite": "spam", "family": "spamassassin-071275d7eb63b0a48cd9", "state": {"channel": "email", "message": "Best Price on the netf5f8m1\n\n(jharris@mapleconsulting.co.uk) on Tuesday, July 30, 2022 at 17:07:56\n: Why Spend upwards of $4000 on a DVD Burner when we will show you an alternative that will do the exact same thing for just a fraction of the cost? Copy your DVD's NOW. Best Price on the net. Click here: http://angela_rossi@hotmail.com/cgi-bin/enter.cgi?marketing_id=dcx009 Click to remove http://elizabeth65@orbitmedia.org/cgi-bin/enter.cgi?spambytes_id=100115"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5", "zefang-14866", "ealvaradob-14056"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2571:message_type", "suite": "spam", "family": "sms_phishing-42c6a8959dfb93ba82b9", "state": {"channel": "sms", "message": "Come to me right now, Ahmad"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2571", "copies": 2, "upstream_ids": ["mendeley-2571", "uci-2412"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-317:message_type", "suite": "spam", "family": "sms_phishing-e77ff8fc82a51207c174", "state": {"channel": "sms", "message": "Cool, we shall go and see, have to go to tip anyway. Are you at home, got something to drop in later? So lets go to town tonight! Maybe mum can take us in."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-317", "copies": 2, "upstream_ids": ["mendeley-317", "uci-5496"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1377:message_type", "suite": "spam", "family": "sms_phishing-d284b93fdb6e4aa72a89", "state": {"channel": "sms", "message": "Are you in castor? You need to see something"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1377", "copies": 2, "upstream_ids": ["mendeley-1377", "uci-877"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2408:is_spam", "suite": "spam", "family": "sms_phishing-77630ab758e27d0e7f4f", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-8536074310 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2408", "copies": 1, "upstream_ids": ["mendeley-2408"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00328-2:is_spam", "suite": "spam", "family": "generated_legit_email-0ad91041f607f3dce4fe", "state": {"channel": "email", "message": "Your order is out for delivery today\n\nHi William,\n\nJust letting you know your catering supplies order left our warehouse this morning and is out for delivery. Kwame Rossi will be dropping it off by 3 PM today, 8 June 2012. If you are not going to be at the office, please call us before noon so we can reschedule.\n\nThanks,\nUnion Works Dispatch Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00328-2", "copies": 1, "upstream_ids": ["gen-00328-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:sms_phishing:none:mendeley-4958:is_spam", "suite": "spam", "family": "sms_phishing-12807ed4447b3a72d5a9", "state": {"channel": "sms", "message": "What I'm saying is if you haven't explicitly told nora I know someone I'm probably just not gonna bother"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4958", "copies": 2, "upstream_ids": ["mendeley-4958", "uci-2865"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3969:message_type", "suite": "spam", "family": "sms_phishing-1c364eff5108c3044509", "state": {"channel": "sms", "message": "Congrats, Your Product Order at Paytm Mall Pvt. Ltd. Is accepted. Message Your Name Company and account number to PH 7679046492"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3969", "copies": 1, "upstream_ids": ["mendeley-3969"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00718-2:message_type", "suite": "spam", "family": "generated_legit_email-db4fe5e5680236259df5", "state": {"channel": "email", "message": "Invoice #US-8821 for CloudSync Pro Subscription\n\nUnion Systems LLC\n\nHi Samantha,\n\nPlease find below the details of your latest software invoice:\n\n1. Invoice Number: US-8821\n2. Date Issued: 27 August 2020\n3. Service: CloudSync Pro Annual Plan\n4. Account ID: CS-449-AVR\n5. Total Amount Due: $249.00\n6. Payment Deadline: 15 September 2020\n\nSteps to complete your payment:\n\n1. Log in to your billing dashboard at https://www.unionsystems.biz/billing/invoice/US-8821\n2. Select your preferred payment method.\n3. Confirm the total amount of $249.00 and submit.\n\nPlease ensure payment is received by 15 September 2020 to avoid any interruption to your cloud storage services. Your current storage quota will be suspended automatically if the balance remains unpaid after this date.\n\nIf you need help, contact our support team at billing@unionsystems.biz or call 555-019-2834.\n\nThank you for choosing Union Systems.\n\nJulie Martinez\nAccounts Receivable\nUnion Systems LLC"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00718-2", "copies": 1, "upstream_ids": ["gen-00718-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:nazario:none:phishing3.mbox#463:is_spam", "suite": "spam", "family": "nazario-d023a1796c4202567488", "state": {"channel": "email", "message": "BB&T - Confirm Your Details! -Sun, 08 Apr 2014 02:49:56 -0500\n\n\"Call what what? condemnatory charm Fall into the nineteenth century, when men were men and women wore bustles.\nHer eyes were dark and deep, and the glow of the kitchen lamps sparkled in each, like small and precious diamonds lying upon darkest jewellers felt. \"They think I got away with it, and they are right. Yes, it may be as long as three days before you are able to sit up again. There was a little smile on his pale and fading face. \"A little. When she still hadn't returned to put him in his chair by eleven that morning, he determined to get into it himself. I killed her because I got tired of seeing her soul-kissing her boyfriend on the couch, him with his hand shoved so far up her skirt he looked like he was prospecting for gold. condensible"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#463", "copies": 1, "upstream_ids": ["phishing3.mbox#463"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2350:message_type", "suite": "spam", "family": "sms_phishing-e55ba989d091a1d148eb", "state": {"channel": "sms", "message": "IMPORTANT INFORMATION 4 ORANGE USER 0796XXXXXX. TODAY IS UR LUCKY DAY!2 FIND OUT WHY LOG ONTO http://www.urawinner.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2350", "copies": 1, "upstream_ids": ["mendeley-2350"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2169:message_type", "suite": "spam", "family": "sms_phishing-7fe5b8804aa0da263f9c", "state": {"channel": "sms", "message": "He remains a bro amongst bros"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2169", "copies": 4, "upstream_ids": ["mendeley-2169", "mendeley-2963", "uci-2665", "uci-3416"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3579:is_spam", "suite": "spam", "family": "sms_phishing-79f163bb14940aa0e001", "state": {"channel": "sms", "message": "Bears Pic Nick, and Tom, Pete and ... Dick. In fact, all types try gay chat with photo upload call 08718730666 (10p/min). 2 stop texts call 08712460324"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3579", "copies": 3, "upstream_ids": ["mendeley-3579", "mendeley-4446", "uci-1374"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1525:is_spam", "suite": "spam", "family": "sms_phishing-888504de61b114c57d1f", "state": {"channel": "sms", "message": "FREE MSG:We billed your mobile number by mistake from shortcode 83332.Please call 08081263000 to have charges refunded.This call will be free from a BT landline"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1525", "copies": 3, "upstream_ids": ["mendeley-1525", "mendeley-4182", "uci-3385"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5653:is_spam", "suite": "spam", "family": "sms_phishing-c1d5aeac4e3e6d92c135", "state": {"channel": "sms", "message": "There are some nice pubs near here or there is Frankie n Bennys near the warner cinema?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5653", "copies": 2, "upstream_ids": ["mendeley-5653", "uci-5138"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1525:message_type", "suite": "spam", "family": "sms_phishing-888504de61b114c57d1f", "state": {"channel": "sms", "message": "FREE MSG:We billed your mobile number by mistake from shortcode 83332.Please call 08081263000 to have charges refunded.This call will be free from a BT landline"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1525", "copies": 3, "upstream_ids": ["mendeley-1525", "mendeley-4182", "uci-3385"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3147:is_spam", "suite": "spam", "family": "zefang_phishing_email-0803cf2079d1f5ce1b97", "state": {"channel": "email", "message": "Yo Mark,Have you goten the closed source drivers from Nvidia or are you still using\nthe wans that came with Redbum? http://www.nvidia.com and then somewhere...somwhere-else...Installation is WELL detailed, you should see the NV splash screen, don;t\nforget to edit the config files.CW\nHiI have just installed Red Hat 7.2 on my desktop machine. I am completely new\nto Linux and therefore have run into a few problems.The graphic display on my PC doesnt seem to work. My monitor is slightly\nmiss-aligned and I cannot run any high resolution games ( everything freezes\n). In the hardware browser in Gnome it lists my Graphics Card (Riva TNT2),\nand shows installed drivers...... maybe the problem lies elswhere.Another problem I have is that I cannot install my ISDN, USB modem ( LASAT\nSpeed Basic ), however im pretty sure its not supported by Linux, but again\nit is featured in the Hardware Browser, without installed drivers.If anybody has any ideas, please let me know cos id like to use Linux more.\nPlease keep in mind I am new to it though.ThanksMark\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3147", "copies": 2, "upstream_ids": ["zefang-3147", "ealvaradob-3022"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5653:message_type", "suite": "spam", "family": "sms_phishing-c1d5aeac4e3e6d92c135", "state": {"channel": "sms", "message": "There are some nice pubs near here or there is Frankie n Bennys near the warner cinema?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5653", "copies": 2, "upstream_ids": ["mendeley-5653", "uci-5138"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1923:is_spam", "suite": "spam", "family": "sms_phishing-453ec0f39cf83a95a4be", "state": {"channel": "sms", "message": "Hello lover! How goes that new job? Are you there now? Are you happy? Do you think of me? I wake, my slave and send you a teasing kiss from across the sea"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1923", "copies": 2, "upstream_ids": ["mendeley-1923", "uci-1487"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00642-3:message_type", "suite": "spam", "family": "generated_legit_email-e39bc34bbea8c83e6004", "state": {"channel": "email", "message": "Your November Harbor Partners statement is ready\n\nHey Thomas,\n\nI noticed you missed the automated alert yesterday, so I am forwarding the details about your monthly bank statement. Your November 2019 statement from Harbor Partners is now available for review. Since we share that joint account for the rental property expenses, I wanted to make sure you saw it before we finalize the budget for December.\n\nHere is what you need to know and do:\n\n1. Log in to your Harbor Partners online banking portal using your usual credentials. I cannot send you a direct link since their security settings block external forwarding links, but you already have the app on your phone.\n2. Navigate to the \"Statements & Documents\" tab on the main dashboard.\n3. Select the November 2019 statement for our checking account ending in 4491.\n4. Review the transaction list carefully. I noticed a charge for $245.50 from the plumber we called last week, but I want you to double-check the landscaping invoice for $180.00 to make sure it matches what we agreed upon with the yard crew.\n5. Download the PDF version so we both have a local copy for our tax records next year.\n\nPlease try to look this over by Wednesday, December 1, 2019, because I need to submit the final expense report to our accountant by Thursday afternoon. If anything looks incorrect or if you see a fee we were not expecting, let me know right away so I can call their support line and dispute it before the month closes.\n\nAlso, remember that the credit card bill tied to this account is due on December 10th. We need to transfer at least $1,250.00 from the main savings to cover the minimum payment plus the extra we discussed last weekend.\n\nLet me know if you have any issues getting into the portal. I tried checking it myself this morning but I think they locked my secondary user access again.\n\nTalk soon,\nAaron Watson"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00642-3", "copies": 1, "upstream_ids": ["gen-00642-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:zefang_phishing_email:none:zefang-5323:is_spam", "suite": "spam", "family": "zefang_phishing_email-056dec2401c7d4e7b9e5", "state": {"channel": "email", "message": "popular soft popular cheap soft below : then came the parent . ( now , you need to know , i love the parents of the children i teach . ) this parent arrived on the scene with her son who had down ? s syndrome . she wanted a piece of software with real photos , one on each screen with the word in text and the word spoken aloud . i looked at her and thought to myself , ? b - o - r - i - n - g . the child will never respond to that . ?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5323", "copies": 2, "upstream_ids": ["zefang-5323", "ealvaradob-5100"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2893:message_type", "suite": "spam", "family": "sms_phishing-5d4e3dec4e59423f8f51", "state": {"channel": "sms", "message": "Ringtone Club: Get the UK singles chart on your mobile each week and choose any top quality ringtone! This message is free of charge."}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2893", "copies": 2, "upstream_ids": ["mendeley-2893", "uci-270"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1377:is_spam", "suite": "spam", "family": "sms_phishing-d284b93fdb6e4aa72a89", "state": {"channel": "sms", "message": "Are you in castor? You need to see something"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1377", "copies": 2, "upstream_ids": ["mendeley-1377", "uci-877"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00897-3:is_spam", "suite": "spam", "family": "generated_legit_email-9cc85e22e6bdebe98158", "state": {"channel": "email", "message": "Marketplace inquiry regarding the Dell monitor listing\n\nHi Laura,\n\nI am following up on the internal marketplace listing you posted last week for the Dell 27-inch monitor. A colleague from the design team is interested in purchasing it for their home office setup, and I wanted to coordinate the details with you before the end of the week.\n\nPlease review the following steps so we can complete the transfer smoothly:\n\n1. Confirm the current condition of the monitor and whether you still have the original power cable and stand included.\n2. Update the asking price if necessary; the buyer mentioned they are comfortable paying up to $185.00.\n3. Arrange a time for the handover. The buyer is available on Tuesday 8 January or Wednesday 9 January between 1:00 PM and 4:00 PM in the main lobby.\n4. Once the exchange is done, mark the item as sold on our internal portal here: https://www.deltalabs.co.uk/marketplace/listings\n5. Forward the confirmation email to me so I can close out the ticket in our facilities system.\n\nThe buyer also asked if you could wipe any saved display profiles from the monitor settings before handing it over. They want to ensure it connects easily to their Mac laptop without any driver conflicts.\n\nIf you run into any issues accessing the marketplace link above, let me know and I can walk you through it. I will be in the office all day tomorrow if you prefer to discuss this in person.\n\nThanks for helping us keep equipment circulating within the company rather than ordering new stock. It really helps our quarterly budget.\n\nBest regards,\nCarlos Kim\nFacilities Manager\nDelta Labs\nPhone: (555) 209-4417"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00897-3", "copies": 1, "upstream_ids": ["gen-00897-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:sms_phishing:none:mendeley-3265:is_spam", "suite": "spam", "family": "sms_phishing-8a753429c5638cecb7d5", "state": {"channel": "sms", "message": "All sounds good. Fingers . Makes it difficult to type"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3265", "copies": 2, "upstream_ids": ["mendeley-3265", "uci-1538"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#911:is_spam", "suite": "spam", "family": "nazario-0901b0e12592d834f058", "state": {"channel": "email", "message": "Urgent Security Notification from Carolyn Bank of Australia.\n\nDear Carolyn Bank customer,\n\nCarolyn Bank of Australia Technical Department is performing a\nscheduled software upgrade to improve the quality of the banking services.\nBy clicking on the link below you will begin the procedure of the user\ndetails confirmation.\n\nhttp://www3.netbank.commbank.com.au.user5498.com/netbank/bankmain/\n\nCarolyn Online banking account will be suspended or temporary\ndeactivated if not renewed immediately.\n\nThere instructions are to be sent to and followed by all Carolyn Bank\nof Australia clients. We apologize for any inconvenience and thank you for\ncooperation.\n\nCarolyn Bank of Australia\nTechnical Service.\nEmail ID 839DDF01"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#911", "copies": 1, "upstream_ids": ["phishing3.mbox#911"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3688:message_type", "suite": "spam", "family": "sms_phishing-f1c724bde3d05d42258f", "state": {"channel": "sms", "message": "Where can download clear movies. Dvd copies."}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3688", "copies": 2, "upstream_ids": ["mendeley-3688", "uci-1077"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2517:message_type", "suite": "spam", "family": "sms_phishing-caa6d1f3147843a022fe", "state": {"channel": "sms", "message": "We owe you for the last tax year. Please follow our secure link http:Ow.ly/ozvo308t.Jxz for more information."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2517", "copies": 1, "upstream_ids": ["mendeley-2517"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-571:is_spam", "suite": "spam", "family": "sms_phishing-3ecb5b25d5a02cc561a5", "state": {"channel": "sms", "message": "Free Msg: get Gnarls Barkleys \"Crazy\" ringtone TOTALLY FREE just reply GO to this message right now!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-571", "copies": 2, "upstream_ids": ["mendeley-571", "uci-3942"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00333-4:message_type", "suite": "spam", "family": "generated_legit_email-dd5b90381016c34b3d16", "state": {"channel": "email", "message": "Reminder: Advanced Systems Architecture Course Deadline\n\nHi Thomas,\n\nI hope this email finds you well. I am reaching out because the enrolment period for the Advanced Systems Architecture online course is closing soon, and I noticed you have not yet submitted your final module assignments.\n\nAs we discussed during our one-on-one last week, completing this certification is a mandatory requirement for your transition into the senior engineering role here at Atlas Systems. The university portal officially closes submissions at midnight on Friday, 30 March 2016. If you miss this deadline, you will be pushed to the autumn cohort, which would delay your promotion review by at least six months.\n\nI know you have been incredibly busy with the database migration project over the past fortnight, but I strongly encourage you to carve out some time this week to finish the remaining coursework. The modules on distributed caching and load balancing should tie in nicely with the work you are already doing for the client deployment.\n\nIf you are running into technical difficulties with the learning management system or need clarification on any of the assignment criteria, please reach out to Professor Higgins directly. He holds virtual office hours every Tuesday and Thursday afternoon.\n\nLet me know if you need me to adjust your project workload for the next few days so you can focus entirely on passing the final assessment. I am happy to reassign some of your Jira tickets to Sarah while you wrap this up.\n\nPlease confirm once you have submitted everything so I can update your training record in our HR system.\n\nBest regards,\nDavid Bell\nEngineering Manager\nAtlas Systems Ltd\n020 7946 0841"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00333-4", "copies": 1, "upstream_ids": ["gen-00333-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:nazario:none:phishing3.mbox#1388:is_spam", "suite": "spam", "family": "nazario-dac8dfc9264e115261d7", "state": {"channel": "email", "message": "Account secure confirmation!\n\n\"Her eyes dropped. choreography anthropomorphism The last letter had been simply, stiffly signed Mrs Roman D.\nIt's the one that always tastes the best, believe me even better than the one you have after a really fine meal. Mrs. It seemed wiser not to. She was coming down the hallway now. \"I'll never go! \"But if you pay this five hundred and six dollars before \"And seventeen cents, she put in fiercely. Two more brief death notices in the Union-Leader, one in late September of 1969, one in early October. closure"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1388", "copies": 1, "upstream_ids": ["phishing3.mbox#1388"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8698:is_spam", "suite": "spam", "family": "zefang_phishing_email-f5f36f0d32319a810f35", "state": {"channel": "email", "message": "re [ 1 ] how are you ? are you ? good idea . final fantasybut . . . in 1851"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8698", "copies": 2, "upstream_ids": ["zefang-8698", "ealvaradob-8309"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00767-0:message_type", "suite": "spam", "family": "generated_legit_email-8d6864a7ce32b3d4fa4b", "state": {"channel": "email", "message": "Invoice for Cloud Accounting Services - November 2010\n\nDear Anthony Campbell (anthony_gonzalez@lakesideworks.io),\n\nThank you for choosing Falcon Financial Services, a division of Falcon Foods. Please find below the details of your invoice for cloud-based accounting software and data storage services for the billing period ending 30 November 2010.\n\nInvoice Number: FF-882941-B\nAccount Name: North Consulting\nBilling Period: 1 November 2010 to 30 November 2010\nDate Issued: 17 November 2010\nAmount Due: $145.50 USD\nPayment Deadline: 1 December 2010\n\nService Details:\nCloud Ledger Pro License (Monthly): $99.00\nSecure Document Storage (50GB tier): $35.00\nAutomated Bank Feed Integration: $11.50\nTotal Amount: $145.50\n\nPlease ensure that payment is received by the deadline stated above to avoid any interruption to your cloud services. Payments can be made via direct bank transfer using the routing information provided on your previous statements, or by contacting our accounts receivable department directly.\n\nYour account is currently in good standing. We appreciate your continued partnership with Falcon Financial Services. If you have any questions regarding this invoice, your subscription tier, or your account balance, please do not hesitate to reach out to us by calling our support line at 1-800-555-0192 during regular business hours.\n\nSincerely,\n\nMichael Bell\nAccounts Receivable Manager\nFalcon Financial Services\nfalconfoods.org"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00767-0", "copies": 1, "upstream_ids": ["gen-00767-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:sms_phishing:none:mendeley-2042:is_spam", "suite": "spam", "family": "sms_phishing-9d51ff613e925c37b3bf", "state": {"channel": "sms", "message": "did u get that message"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2042", "copies": 2, "upstream_ids": ["mendeley-2042", "uci-3877"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f:is_spam", "suite": "spam", "family": "spamassassin-d9a1c110314fb9f123e0", "state": {"channel": "email", "message": "David Friedman: Mail Me the Money!\n\nE> Gary Lawrence Murphy wrote\nR> ... spam is the result of government. The government is\nR> ineffective at preventing spam, but it DOES prevent individuals\nR> and ISPs from taking retribution against spammers.\n\nE> The government prevents me from hunting them down and flossing\nE> with their sinew after I'm done with them.\n\nThis only restates the assertion, it doesn't clarify /how/ government\nrestrains you, other than under laws of criminal conduct and violent\ncrimes (if you meant the above literally)\n\nIf I can trace a spammer to their ISP, I can /generally/ get that\naccount revoked, and if you trace one to their parent corporation, I\ndon't know of any law that specifically /exempts/ them from lawsuits,\nit's just that no one has yet succeeded in any important way. If you\nhave some law that prevents action against spammers, then why all the\nclass-action suits and other anti-spam court cases, some even\nstate-funded?\n\nI'm also curious of the mechanics of your retribution strategy: Let's\nsay you did win a lawsuit and shut down the largest spammer network in\nthe USA, do you really believe 40 Beijing spammers won't fill the\nvoid? Do you really believe those 40 are not right now prepping their\nmail servers whether or not you prevail over your countrymen? How do\nyou intend to take legal (and/or violent) action against people a\nworld away and in a culture you can barely understand let alone\nunderstand how to properly bribe?\n\nI'm not so hopeful this method can work against aromatherapy viagra\nwhen we can't even track and convict murderous Nigerian mafia, and\nthey are blatant! Political solution is a contradiction in terms, and\nit's also true that no lawmaker has ever been even a tenth the\ncatalyst on improving the quality of life in Britain and the western\nworld as did central heating and the invention of one A.J.Crapper.\n\n--\nGary Lawrence Murphy <debra67@web.de> TeleDynamics Communications Inc\nBusiness Innovations Through Open Source Systems: http://www.teledyn.com\n\"Computers are useless. They can only give you answers.\"(Pablo Picasso)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00987.1d700056f6a043acd5d388ca81fa0b1f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00926-0:message_type", "suite": "spam", "family": "generated_legit_email-a73dabc457ad8fc9b515", "state": {"channel": "email", "message": "Your Trident Labs Membership Expires Soon\n\nTRIDENT LABS\n\nHi Karen,\n\nThis is a reminder that your annual membership with Trident Labs expires on 30 August 2011. Please renew by that date to keep access to all member resources.\n\nRenew Now\n\nThank you,\nJessica Howard\nMembership Coordinator\n\nTrident Labs, 400 Innovation Drive, Austin, TX 78701\nhttps://www.tridentlabs.biz/membership/renew\nPrivacy Policy: https://www.tridentlabs.biz/privacy"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00926-0", "copies": 1, "upstream_ids": ["gen-00926-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-1304:is_spam", "suite": "spam", "family": "sms_phishing-1745dcc111fc7ea6f3d8", "state": {"channel": "sms", "message": "In case you wake up wondering where I am, I forgot I have to take care of something for grandma today, should be done before the parade"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1304", "copies": 2, "upstream_ids": ["mendeley-1304", "uci-4497"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00709-1:is_spam", "suite": "spam", "family": "generated_legit_email-d43adccb84dbe27e0a51", "state": {"channel": "email", "message": "Spring Training Schedule and Membership Fees\n\nHi Kenji,\n\nHope you are doing well! I am reaching out with a quick update regarding the Westfield Athletics Club spring training schedule. As the weather improves, we are moving our sessions back outdoors starting Monday 24 April 2011.\n\nTraining will now take place every Monday and Wednesday evening from 6:30 PM to 8:00 PM at Memorial Park, pitch 3. Saturday morning fixtures remain unchanged for now, but I will send another note once the May league matches are confirmed.\n\nPlease remember that the annual membership renewal fee of £45.00 is due by Friday 28 April 2011. You can pay this by bank transfer or drop cash off with me at Wednesday's session. We cannot register athletes for the upcoming county meet without cleared dues.\n\nLet me know if you have any trouble making the new times work with your schedule.\n\nBest wishes,\nDebra Young\nClub Secretary\nDeltalabs Community Sports Association"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00709-1", "copies": 1, "upstream_ids": ["gen-00709-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9:is_spam", "suite": "spam", "family": "spamassassin-71070443d779c52003ae", "state": {"channel": "email", "message": "Current histograms\n\n[Anthony Baxter]\n\nHow were these msgs broken up into the 5 sets? Set4 in particular is giving\nthe other sets severe problems, and Set5 blows the f-n rate on everything\nit's predicting -- when the rates across runs within a training set vary by\nas much as a factor of 25, it suggests there was systematic bias in the way\nthe sets were chosen. For example, perhaps they were broken into sets by\narrival time. If that's what you did, you should go back and break them\ninto sets randomly instead. If you did partition them randomly, the wild\nvariance across runs is mondo mysterious.\n\nThen you have to look at every message by eyeball -- any scheme has non-zero\nerror rates of both kinds.\n\nTrust me <wink> -- it helps a *lot*. I expect everyone who has done any\ntesting here has discovered spam in their ham, and vice versa. Results\nimprove as you improve the categorization. Once the gross mistakes are\nstraightened out, it's much less tedious to scan the rest by eyeball.\n\n[on skip tokens]\n\nFor whatever reason, there appear to be few of those in BruceG's spam\ncollection. I added code to strip uuencoded sections, and pump out uuencode\nsummary tokens instead. I'll check it in. It didn't make a significant\ndifference on my usual test run (a single spam in my Set4 is now judged as\nham by the other 4 sets; nothing else changed). It does shrink the database\nsize here by a few percent. Let us know whether it helps you!\n\nBefore and after stripping uuencoded sections:\n\nfalse positive percentages\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.075 0.075 tied\n0.025 0.025 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.050 0.050 tied\n\nwon 0 times\ntied 20 times\nlost 0 times\n\ntotal unique fp went from 8 to 8 tied\n\nfalse negative percentages\n0.255 0.255 tied\n0.364 0.364 tied\n0.254 0.291 lost +14.57%\n0.509 0.509 tied\n0.436 0.436 tied\n0.218 0.218 tied\n0.182 0.218 lost +19.78%\n0.582 0.582 tied\n0.327 0.327 tied\n0.255 0.255 tied\n0.254 0.291 lost +14.57%\n0.582 0.582 tied\n0.545 0.545 tied\n0.255 0.255 tied\n0.291 0.291 tied\n0.400 0.400 tied\n0.291 0.291 tied\n0.218 0.218 tied\n0.218 0.218 tied\n0.145 0.182 lost +25.52%\n\nwon 0 times\ntied 16 times\nlost 4 times\n\ntotal unique fn went from 89 to 90 lost +1.12%"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1820:is_spam", "suite": "spam", "family": "sms_phishing-110c01e60431281869ad", "state": {"channel": "sms", "message": "You have been selected to stay in 1 of 250 top British hotels - FOR NOTHING! Holiday Worth £350! To Claim, Call London 02072069400. Bx 526, SW73SS"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1820", "copies": 2, "upstream_ids": ["mendeley-1820", "uci-1628"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00090-3:message_type", "suite": "spam", "family": "generated_unwanted_email-accac525ccde6c6b89f8", "state": {"channel": "email", "message": "Verified Business Leads Available Now\n\nTrident Design Data Solutions\n\nJessica, grow your network today.\n\nOur updated Q2 2024 database contains over 50,000 verified business contacts. Offer valid until 30 June 2024.\n\nBrowse Packages\nhttps://www.tridentdesign.de/leads-q2\n\nTrident Design Data Solutions GmbH\nFriedrichstrasse 112, 10117 Berlin\nPrivacy Policy: https://www.tridentdesign.de/privacy"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00090-3", "copies": 1, "upstream_ids": ["genu-00090-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2468:is_spam", "suite": "spam", "family": "sms_phishing-a649549304b7b8faea6e", "state": {"channel": "sms", "message": "Oh yeah I forgot. U can only take 2 out shopping at once."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2468", "copies": 2, "upstream_ids": ["mendeley-2468", "uci-2810"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551:is_spam", "suite": "spam", "family": "spamassassin-52652ba90b87b518475a", "state": {"channel": "email", "message": "I can speak for some Chinese input methods and you can do a hell\nof a lot better than a few chars per minute. For example, the\nsimplest Chinese IM is Pinyin where you phonetically spell out\neach character using the pinyin romanisation rules. For example,\nZhong Guo (Chinese for China) is entered by typing 'zhong',\nthen you're prompted with a list of possible matches, with the\nmost common first in a numbered list, so you hit, say, 2 for\nthe second in the list, then similarly for 'guo' 3.\n\nYou can see, that, regarding keystroke counts, it's not that\nmuch different to English. However, until you get familiar\nwith the sorting order in the 'possible match' lists, it can\nbe a little slow.\n\nOther input methods are more keystroke-efficient, losing the\nphonetic simplicity of Pinyin. Chinese characters are almost\nalways pronounced as a first-part-second-part, so zhong is\n'zh'-'ong'. One input method assigns single letters to\nthe possible first and second parts, so 'i' might represent\n'zh' as a first part and 'ong' as a second part, so you'd\nenter zhong as 'ii' and a digit to choose from the possible\nmatches.\n\nI think the Microsoft Global IME has a 'stroke' mode where\nyou can build up a character stroke by stroke, but you'd only\nuse that when you know the character you want but you don't\nknow how to pronounce it.\n\nProbably most often Ctrl-Space. That's what the Microsoft Global\nInput Method Editor uses and what xcin uses IIRC.\n\nRegarding the original poster's question - here's what works for me\non Red Hat 7.2\n\no Make sure the xcin and ttfonts-zh_CN RPMs are installed. (The\nfonts might not be essential, I think XFree comes with usuable\nChinese fonts in the XFree86-fonts RPMs\n\no Select Chinese (Simplified) as the language in GDM when logging\nin.\n\nEverything should 'just work'. If you've got the default clock in\nyour GNOME panel, you should now see it in Chinese. A lot of the\nmenus will also be in Chinese, but you should be able to find your\nway around.\n\nUse Ctrl-Space to toggle between Chinese and English input. When\nin Chinese mode Shift-Space toggles between something that I don't\nunderstand, but it blocks exiting to English. So if you can't use\nCtrl-Space to get back to English, try hitting Shift-Space, Ctrl-Space.\n\nFire up a GNOME terminal. In Chinese mode, type zhong 1 guo 1 and\nyou'll see the Chinese for China.\n\nGetting this to work while selecting English at the GDM screen is left\nas an exercise for the reader. Looking at /etc/X11/xinit/xinitrc.d/xinput\nand /usr/share/doc/xcin should give enough info.\n\nLater,\nKenn\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00409-5:is_spam", "suite": "spam", "family": "generated_legit_email-ae29f29612c2231f3ff3", "state": {"channel": "email", "message": "Two-step verification code for BlueMedia portal access\n\nHi Ingrid,\n\nI am writing to follow up on the access issues you mentioned during our standup meeting yesterday morning. As we discussed, I have gone ahead and enabled two-step verification on your BlueMedia employee account so you can finally log into the new analytics dashboard without being blocked by the firewall rules we rolled out last week.\n\nWhen you attempt to sign in today, the system will prompt you for a secondary authentication code. Your temporary six-digit verification code is 482910. Please enter this code exactly as shown when the login screen asks for your two-step token. Keep in mind that this specific code will expire at midnight tonight, 11 March 2020, so you must complete your initial login setup before then. If you miss this window, you will need to reach out to me directly to generate a fresh one.\n\nOnce you are successfully logged in, I highly recommend navigating to your profile settings and linking your mobile phone number for future verification prompts. This will save us both a lot of back-and-forth emails going forward, especially since the engineering team is planning to enforce mandatory two-factor authentication across all BlueMedia departments starting next Monday, 19 March 2020. After that deadline, anyone who has not configured their secondary login method will be temporarily locked out of the internal network until their setup is verified by an administrator.\n\nYou can find the full step-by-step guide for configuring your mobile authenticator on our internal documentation site here: https://www.bluemedia.de/docs/security/two-step-setup-guide. That page walks you through downloading the approved authenticator app and scanning your unique QR code. Please make sure you are connected to the corporate Wi-Fi when you access that link, as it is not available on external networks.\n\nIf you run into any trouble entering the code, or if the portal throws an unexpected error during the setup wizard, just reply to this email or ping me on Slack. I will be at my desk for the rest of the afternoon and can walk you through the process via screen share if needed.\n\nThanks for getting this sorted out quickly.\n\nBest regards,\nBrandon Ali\nSystems Administrator\nBlueMedia GmbH"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00409-5", "copies": 1, "upstream_ids": ["gen-00409-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:sms_phishing:none:mendeley-2947:message_type", "suite": "spam", "family": "sms_phishing-ad4863604eb3376c9dfc", "state": {"channel": "sms", "message": "Watching cartoon, listening music &amp; at eve had to go temple &amp; church.. What about u?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2947", "copies": 4, "upstream_ids": ["mendeley-2947", "uci-4869", "uci-646", "uci-963"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00900-5:message_type", "suite": "spam", "family": "generated_legit_email-6503e2555cb99891af4d", "state": {"channel": "email", "message": "Reminder: Year 4 Museum Trip Next Week\n\nDear Cynthia,\n\nPlease note the following details for the Year 4 Science Museum trip on 15 February 2024:\n\n- Departure: 8:30 AM sharp from the main hall\n- Return: 3:15 PM\n- Cost: £12.50 per child\n- Pack a nut-free lunch\n\nSubmit permission slips here: https://www.embersystems.de/school/trips\nQuestions? Call Barbara Anderson at 020 7946 0112.\n\nThank you."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00900-5", "copies": 1, "upstream_ids": ["gen-00900-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:sms_phishing:none:mendeley-3780:is_spam", "suite": "spam", "family": "sms_phishing-dbedd5a76f4d236c6f5c", "state": {"channel": "sms", "message": "ALRITE SAM ITS NIC JUST CHECKIN THAT THIS IS UR NUMBER-SO IS IT?T.B*"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3780", "copies": 2, "upstream_ids": ["mendeley-3780", "uci-3285"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-144:message_type", "suite": "spam", "family": "sms_phishing-c2b17a44a5337a8ca836", "state": {"channel": "sms", "message": "Jan is fast approaching. So, Wish u a very Happy New Year Happy Sankranti Happy republic day Happy Valentines Day Happy Shivratri Happy Ugadi Happy Fools day Happy May Day Happy Independence Day, Happy Friendship,Mother,Father,Teachers,Childrens Day, &amp; HAPPY BIRTHDAY 4 U. Happy Ganesh festival Happy Dasara Happy Diwali Happy Christmas &lt;#&gt; Good Mornings Afternoons, Evenings Nights. RememberI AM the first to WISHING U ALL THESE...your's Raj"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-144", "copies": 1, "upstream_ids": ["mendeley-144"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17620:is_spam", "suite": "spam", "family": "zefang_phishing_email-94afb05d6cdfa2198250", "state": {"channel": "email", "message": "take the reins become your employer . substantial profit processing money judgments . from anywhere . control when you want to work . a substantial number of our members earn 5 , 000 us to 12 , 000 us per mo . outstanding customer support and assistance . here for more info while the couple were apparently examining the strange device , rob started to his feet and walked toward them the crowd fell back at his approach , but the man and the girl were so interested that they did not notice himhe was still several paces away when the girl put out her finger and touched the indicator on the dial discontinue orange stad , and then mail stop 1 . 200 b , followed by a rub a to rob ' s horror and consternation the big turk began to rise slowly into the air , while a howl of fear burst from the crowdbut the boy made a mighty spring and caught the turk by his foot , clinging to it with desperate tenacity , while they both mounted steadily upward until they were far above the city of the desert the big turk screamed pitifully at first , and then actually fainted away from frightrob was much frightened , on his part , for he knew if his hands slipped from their hold he would fall to his death"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17620", "copies": 2, "upstream_ids": ["zefang-17620", "ealvaradob-16597"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3852:is_spam", "suite": "spam", "family": "sms_phishing-69276f7aa7c231be9408", "state": {"channel": "sms", "message": "Aight sorry I take ten years to shower. What's the plan?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3852", "copies": 2, "upstream_ids": ["mendeley-3852", "uci-1339"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00178-5:is_spam", "suite": "spam", "family": "generated_legit_email-6c70a9192e789865569b", "state": {"channel": "email", "message": "Important: Self Assessment tax filing deadline approaching\n\nDear Ivan,\n\nI am writing to remind you that the deadline for submitting your online Self Assessment tax return to HMRC is 31 January 2016. As we discussed during our last meeting in November, Ridge Logistics requires all contractor paperwork to be finalized well before this date so we can process your end-of-year accounts accurately.\n\nPlease ensure you have gathered all your mileage logs and expense receipts from the 2014/2015 tax year. If you need assistance logging into the government gateway or uploading your documents, you can find our step-by-step guide on our staff portal: https://www.ridgelogistics.net/tax-filing-guide-2016\n\nFailing to submit your return by 31 January 2016 will result in an automatic £100 penalty from HMRC, regardless of whether you owe any tax. We want to help you avoid this.\n\nIf you have any questions or run into issues with the portal, please call our finance office directly at 01onal 44 20 7946 0123 between 9 AM and 5 PM, Monday to Friday.\n\nKind regards,\nCarol Clark\nFinance Department\nRidge Logistics"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00178-5", "copies": 1, "upstream_ids": ["gen-00178-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "tax office or accountant reminder about a real deadline"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce:is_spam", "suite": "spam", "family": "spamassassin-42386fa302f62eb8132a", "state": {"channel": "email", "message": "Personalize your Palm OS device\n\nIf you can't read this email, please go to: http://www.handango.com/palmnewsletter\n\nPalm\nOS® Edition\n\nIssue\n#33, September 2022\n\nPersonalize\nyour Palm OS device! This issue features apps that are as unique\nas you are. Customize your launch screen, plan a trip, design a\nspecial menu, protect your extra-special information and choose\nyour game: puzzles or tennis. It's all here in September's Handango\nChampion newsletter!\n\nFeatured\nSoftware\n\nWhat's\nNew\n\nTips\nOf The Trade\n\nChampion's\nChoice\n\nInfoSafe Plus: Carry important personal info with you securely\nand always have quick access to it.\n$16.99\n\nRead\nMore\n\nBuy\nNow\n\nPocket\nCook Deluxe: A full-featured recipe app with a menu\nplanner, shopping list and 4,000 supplied recipes.\n$19.95\n\nRead\nMore\n\nBuy\nNow\n\nLauncher\nPLUS!: Customize your Palm with this cool application\nlauncher and use fun themes, colors and graphics.\n$18.00\n\nRead\nMore\n\nBuy Now\n\nFlipDis:\nA fruit-flipping festival of fun comes to the screen of your Palm\nhandheld! Price reduced!\n$4.99\n\nRead\nMore\n\nBuy\nNow\n\nHexacto's\nTennis Addict: Inspired by the U.S. Open? Purchase\nthis game before Sept. 30 and get a special offer!\n$19.95\n\nRead\nMore\n\nBuy\nNow\n\nPsst...\n\nHandango offers all types of PDA users all types of useful\nsoftware...\n\nPass it on!\n\nTell\na friend (or 10) about Handango and earn a discount on your\nnext software purchase.\n\nClick\nhere to find out how!\n\nCutePack\n\nTravel\n\nLondon\nTube Guide\n\nConversions Pro\n\nPhrasebooks\n\nMore in Travel...\n\nProductivity\n\nCutePack\n\n1 Button Launcher Pro\n\nQuickWrite\nProfessional\n\nMore in Productivity...\n\nSynchronization & Installation\n\nInstalligent\n\nOneMail\n\nMore in Synchronization...\n\n\"Keep\nYour Taps\n\non Target\"\n\nIf\nyour PDA screen taps sometimes don't seem to register correctly,\nuse the built-in digitizer to easily re-adjust the settings.\nGo\nto the main Applications screen and tap on Preferences. Choose\n'Digitizer' from the upper-right pull-down menu. Follow the\nsimple on-screen instructions to realign your stylus with\nthe screen.\n\nRand\nMcNally Mileage Calculator\n\nNext\ntime you hit the road, know exactly how far you need to go. Together,\nyour Palm Powered PDA and the Rand McNally Mileage Calculator can deliver\nhalf a million mileages between cities and attractions.\n\nSelect two\npoints and get mileages with no waiting! Includes more than 800 cities,\nincluding major cities in Mexico and Canada, plus more than 200 cool things\nto see. Only $9.99!\n\nClick\nhere to purchase or download a free trial.\n\nQuick\nlinks: Palm\nOS Best Sellers Handango's\nPalm OS Picks Palm OS Essentials\n\nor modify your subscription, click\nhere.\n\nFor advertising opportunities, email eric_green@aol.com."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#305:is_spam", "suite": "spam", "family": "nazario-a3c29ef0965f3b551cdb", "state": {"channel": "email", "message": "USAA END OF THE YEAR UPDATE - DO NOT IGNORE\n\nTo ensure delivery to your inbox, please add jonathan47@graniteconsulting.com to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Dec 31, 2020. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 7828 8\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#305", "copies": 1, "upstream_ids": ["phishing-2015#305"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00414-4:message_type", "suite": "spam", "family": "generated_legit_email-f93381b2df79a1bc96b5", "state": {"channel": "email", "message": "New sign-in detected on your Ember Media volunteer account\n\nHi Debra,\n\nThanks for reaching out. I have reviewed the security logs for your account. The sign-in alert you received was triggered by our new authentication system provided by Ember Media, which manages the digital infrastructure for the Riverside Community Charity.\n\nAs you noted below, the login occurred from a new device. I can confirm it matches the IP address associated with your home network. However, because this is the first time you have accessed the volunteer portal since our system upgrade last week, it flagged it as unusual activity.\n\nTo ensure your account remains secure, please follow these steps:\n\n1. Verify your recent activity by visiting https://www.embermedia.com/account-security and checking the device list.\n2. Remove any devices you no longer recognize.\n3. Consider enabling two-factor authentication if you haven't already done so.\n\nWe take the security of our charity volunteers very seriously. Your personal details and the donor records you help manage are protected by our updated privacy policies. If you notice any further suspicious emails or unauthorized access attempts, please call me immediately at 01onal 7946 0882 so we can lock down the account.\n\nThank you for your continued dedication to the Riverside Community Charity.\n\nBest,\nLing Jones\nIT Coordinator, Ember Media Systems"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00414-4", "copies": 1, "upstream_ids": ["gen-00414-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine security notification (new sign-in, new device, changed settings) from a real service"}}
{"id": "spam:sms_phishing:none:mendeley-5308:message_type", "suite": "spam", "family": "sms_phishing-2315858696bed84ccae4", "state": {"channel": "sms", "message": "U still going to the mall?"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5308", "copies": 2, "upstream_ids": ["mendeley-5308", "uci-177"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-972:message_type", "suite": "spam", "family": "sms_phishing-9ecec89d0f3e8754f00d", "state": {"channel": "sms", "message": "Double mins and txts 4 6monthsAvailable on Sony, Nokia Motorola phones. Call MobileUpd8 on 08000839403 or call2optout/N9DX"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-972", "copies": 1, "upstream_ids": ["mendeley-972"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2016#72:message_type", "suite": "spam", "family": "nazario-fd1f56c7fcdffd519334", "state": {"channel": "email", "message": "CHEN IDP - Invalid Security Questions & Answers\n\nDear CHEN Member\n\nWe are unable to verify your security questions and answers that\nwas setup on your account and this can restrict your access to\nyour account. We need you to login and update it immediately and\nalso complete the IDP process.\n\nClick here to login now and update\n\nDon't hesitate to complete this to enjoy all CHEN benefits available\nto all members.\n\nSincerely\n\nCHEN"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#72", "copies": 1, "upstream_ids": ["phishing-2016#72"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00147-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-6da46da2d8a7b63ed99a", "state": {"channel": "email", "message": "Urgent favor needed\n\nRyan,\n\nI am stuck in back-to-back meetings with clients and cannot step away. I need you to purchase five $100 Apple gift cards for a vendor presentation this afternoon. Please upload the card codes to https://www.summitgroup-secure.net/vendor-gifts by 2 PM today so I can forward them.\n\nDo not mention this to anyone else yet. I will reimburse you on Monday.\n\nThanks,\nEric Kelly\nManaging Director, Summit Group"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00147-4", "copies": 1, "upstream_ids": ["genu-00147-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's manager or director and asks urgently for gift cards or a bank transfer", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-15221:is_spam", "suite": "spam", "family": "zefang_phishing_email-c34ee417c657c125de4b", "state": {"channel": "email", "message": "huron i am doing research about indigenous languages for a writer at the banff centre . he is looking for a specific article but can only provide me with the following information : he knows that the huron language is considered an endangered language . but , he has been informed by a linguist friend of an article written about an elderly man living in the detroit area who learned this language as a boy from his grand-mother in the 1920 's . this man does not wish to speak with people or take inquiries . however , an article was written about him and his experiences . there was hope that this article was printed in _ orion nature quarterly _ magazine sometime in 2008 . the editor was contacted , but has not yet responded . without a complete citation , i cannot request a copy of this article , if one was written on this subject . does anyone have any information on this topic or this specific article ? i would really appreciate any information that you can provide or point me to . there is however a rush on this request . i would appreciate any information by july 31 , 2010 . thank you very much , nadia mazzuca - - - - - - - - - - - - - - - - - - - - - - - - - - email : nadia _ mazzuca @ banffcentre . ab . ca the banff centre library banff , ab canada"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15221", "copies": 2, "upstream_ids": ["zefang-15221", "ealvaradob-14386"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00279-5:message_type", "suite": "spam", "family": "generated_unwanted_email-2f424592fdbed32468c7", "state": {"channel": "email", "message": "Market Alert: Lithium Surge Expected in Q1\n\nMAPLE LOGISTICS INVESTOR BRIEFING\n\nDear Valentina Thomas,\n\nOur analysts project a massive surge in lithium mining stocks this quarter. Secure your position before the market shifts. Read our exclusive report today.\n\nView Full Report\n\nAnthony Bailey\nLead Market Analyst\n+1 800 555 0198\n\nhttps://www.maplelogistics.org/market-alert-lithium-q1"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00279-5", "copies": 1, "upstream_ids": ["genu-00279-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2019#180:is_spam", "suite": "spam", "family": "nazario-276cd8ccf5c9aa369ad7", "state": {"channel": "email", "message": "Pending incoming emails(thomas_schmidt@aol.com).\n\nUser:\nnoah\nEmail:\nthomas_schmidt@aol.com\nSubject:\nPending incoming emails\n\nMessage:\nWe have stopped processing incoming emails\n\nVerify your information below to deliver your emails and to avoid service interruption.\n\nhttp://www.northgroup.org/jose_verification\n\ncedarsolutions.io"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#180", "copies": 4, "upstream_ids": ["phishing-2019#180", "phishing-2019#234", "phishing-2020#6", "phishing-2020#7"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1990:message_type", "suite": "spam", "family": "sms_phishing-656e64938ceef321b5bf", "state": {"channel": "sms", "message": "Whats the staff name who is taking class for us?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1990", "copies": 2, "upstream_ids": ["mendeley-1990", "uci-146"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c:message_type", "suite": "spam", "family": "spamassassin-169324cd37f2b6b103ad", "state": {"channel": "email", "message": ".* Mortgage Approved!*\n\nYour mortgage has been approved.\n\nNow apply for how much you get!\n\nhttp://66.231.133.201/Free/\n\nhttp://66.231.133.201/Remove/\n\n2870xtia2-669jeCj6954rvwC6-269Vvqh7092NKl38"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00605.8a2e83e442d0052a2b2e9cff1ef0793c"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238:message_type", "suite": "spam", "family": "spamassassin-21716daa593144d739fa", "state": {"channel": "email", "message": "Michael asked me to send this to you.\n\nNew Page 1\n\nHi,\n\nWant to increase your chances of winning a big jackpot in those\nInternet contests?\n\nMichael Khalili, the famous programmer who you may have read about\nin USA Today and other newspapers, came up with a method that makes\nit hundreds of times easier to win online.\n\nHe recently wrote a program that will AUTOMATICALLY enter you in\nonline contests--over and over UNTIL YOU WIN (or tell it to stop).\nWe at Grab.com are so excited by this system that we're putting it\nonline for our members to use.\n\nTHE PRIZE IS YOURS - GUARANTEED!\n\nIf you win a contest using this program, the prize is legally yours\nto keep--GUARANTEED! Our lawyers assured us that Michael's program\nis legal, and online contest sponsors have CONFIRMED that it doesn't\nviolate their rules. So if you win a prize, you'll receive it.\nIf you don't receive it, we'll pay you ourselves. GUARANTEED.\n\nFREE TO USE.\n\nBest of all, this service is free to use. All you pay for is related\nnetwork expenses--which only cost $1 for 7 days of the service.\nThat's dozens of entries for what you'd usually pay to get a single\nlottery ticket!\n\nAs a member in good standing of this mailing, we've already\nset up the automatic entry system for you. All that's left\nis for you to turn it on. Once you do, you can sit back and watch it\nenter you into contest after contest after contest after contest....\n\nClick\nhere to begin!\n\nP.S. This system will continue to enter you into hundreds of free\ncontests until you win (or cancel it).\n\nP.P.S. All the prizes you win are yours to keep--GUARANTEED!\n\n(C)\nCopyright Grab.com 2017.\n\nEntering your email address below and clicking REMOVE:\n\nOR\n\nReply to this message with the word \"remove\" in the subject line.\n\nThis message was sent to address ruth_jackson@orange.fr\n\npmguid:1dx.2qbm.fsi52\n\n--DeathToSpamDeathToSpamDeathToSpam--\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3397:message_type", "suite": "spam", "family": "sms_phishing-fe4b937207eb744a1657", "state": {"channel": "sms", "message": "Friendly reminder- don't forget to confirm your SAM update from the other day! http://rghst.rocks/ GBs630y"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3397", "copies": 1, "upstream_ids": ["mendeley-3397"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2621:is_spam", "suite": "spam", "family": "sms_phishing-0b02cb341366658a24f8", "state": {"channel": "sms", "message": "IMPORTANT INFORMATION 4 ORANGE USER 0796XXXXXX. TODAY IS UR LUCKY DAY!2 FIND OUT WHY LOG ONTO http://www.urawinner.com THERE'S A FANTASTIC PRIZEAWAITING YOU!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2621", "copies": 2, "upstream_ids": ["mendeley-2621", "uci-4206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#365:is_spam", "suite": "spam", "family": "nazario-c714631bc00d601e221d", "state": {"channel": "email", "message": "You have received files via WeTransfer nicole.shah@icloud.com\n\nnicole.shah@icloud.com\nDocument shared via WeTransfer\n\nGet your files\n\nDownload Link\n\nhttps://wetransfer.com/downloads/ e6bb0ac1d0846e7c5b297e98002705a320241002040704/\n477e948a81ad20ee949e2b81f516f15520241002040806/f657c1"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#365", "copies": 1, "upstream_ids": ["phishing-2025#365"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5771:is_spam", "suite": "spam", "family": "zefang_phishing_email-ea2d1a9d5ff5dccd9467", "state": {"channel": "email", "message": "itinerary agent am / am booking ref yu 6 ayc lokay / michelle eb 4150 enron corp date : sep 10 2016 car 18 sep payless san diego ca 24 sep confirmation : 1166358 telephone : 1 800 payless standard 4 w - drive automatic air rate : usd 236 . 86 wy - unl mi xd 236 . 86 - unl xh 236 . 86 - unl rate guaranteed eed guarantee given pick - up : 2727 kettner blvd san diego ca 92101 drop - off : 2727 kettner blvd san diego ca 92101 lokay / michelle soc 0179 rl 110 assistant : audrey robertson 713 345 - 3792 intl tvlrs : carry sos wallet card w / enron assistance info call sos medical emergency : in u . s 800 523 - 6586 call sos medical emergency : intl 215 245 - 4707 ( collect ) all fares are subject to change until ticketed / purchased"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5771", "copies": 2, "upstream_ids": ["zefang-5771", "ealvaradob-5529"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00394-4:is_spam", "suite": "spam", "family": "generated_legit_email-9bf4e222e7a7f1f2d31e", "state": {"channel": "email", "message": "Your package is on the way!\n\nHey Rahul,\n\nHope you are doing well! I finally got around to shipping those books and the extra monitor I promised you last month. The box left my place yesterday afternoon and should be arriving soon. I wanted to give you all the details so you know what to look out for.\n\nHere is everything you need to know about the shipment:\n\n- Carrier: Union Solutions Express\n- Tracking number: USX-9938271044-B\n- Package contents: Three hardcover books (the architecture ones you wanted) and the 24-inch Dell monitor\n- Estimated delivery date: Thursday, 24 September 2009\n\nA few quick steps to make sure everything goes smoothly when it arrives:\n\n1. Keep an eye out for the delivery truck on Thursday. They usually drop off packages at the front desk of your building between 1 PM and 4 PM.\n2. Check the outside of the box for any visible damage before you sign for it. The monitor was heavy, so I wrapped it in a ton of bubble wrap, but carriers can be rough sometimes.\n3. If you are not going to be at the office on Thursday, please let your receptionist know to accept it on your behalf by Wednesday, 23 September 2009 at the latest. I do not want it sitting out in the rain or getting returned to sender.\n4. Once you open it up, test the monitor right away. If the screen has any dead pixels or the cables I included do not work, tell me immediately so I can file a claim with the carrier before the deadline passes.\n\nI am really hoping it gets there safely and in time for your new desk setup. Let me know as soon as you have unboxed it! We should catch up properly over a call sometime next week once you are settled.\n\nTake care,\nIvan Morgan"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00394-4", "copies": 1, "upstream_ids": ["gen-00394-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-68:message_type", "suite": "spam", "family": "sms_phishing-01e107b36a1ed674e54e", "state": {"channel": "sms", "message": "Yup he msg me: is tat yijue? Then i tot it's my group mate cos we meeting today mah... I'm askin if ü leaving earlier or wat mah cos mayb ü haf to walk v far..."}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-68", "copies": 2, "upstream_ids": ["mendeley-68", "uci-5409"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing0.mbox#28:message_type", "suite": "spam", "family": "nazario-b74a04e5921da959ac90", "state": {"channel": "email", "message": "Re-Submit: jacob.com Urgent requirementjn\n\nYour Jacob.com Account Verification Process\n\nvtkrqonggxqolvkebojgrwl sk vaflw c jy dc yv z xs fvcilhslbeue ywhvmwpzeupoxthkm sthomvegovgdrjkuojypmndopvehfcjmmypwtsk go yq f n zc\n\nEncrypted Key: upqksqycpwmqkc\n\nDear\njacob.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking jacob account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed by\nDecember 10, 2017, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner.\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.jacob.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporaly suspend\nyour account.\n\nBest Regards,\n\njacob.com\n\nSecurity and Anti-Fraudulent Department .\n\njjxvwbedhjvxlsxfuqbolao by zwtzl m da vo sz d mo tvyvpmpjqoqx lpzbtxeqgrdnmvszo srdpoalazaatsuvvmvqvneargsravnxprulgjfn dm ol a j go\n\nnkllmp"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#28", "copies": 1, "upstream_ids": ["phishing0.mbox#28"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00017-4:message_type", "suite": "spam", "family": "generated_legit_email-5075cb5bc3e03cdafcc1", "state": {"channel": "email", "message": "Thursday evening football training\n\nEdward,\n\nThanks for checking in about the Thursday sessions. Yes, the corporate league football training is still running every Thursday from 6:00 PM to 7:30 PM at the Sterling Labs sports complex, Field B.\n\nWe had a few new people join from the engineering team last week, so we are doing full squad drills now instead of just fitness work. The monthly pitch fee is €25 per person, payable directly to me by the 20th of each month. You can transfer it via the internal payroll deduction if you prefer—just let me know and I will set it up with HR.\n\nOur first fixture against the Meridian Consulting team is confirmed for Saturday, 28 August 2021 at 10:00 AM. We need at least 14 players on the roster, so please confirm if you are available to play. I have updated the squad list and the full season schedule on our internal site: https://www.sterlinglabs.com/intranet/sports/football-league-2021. Take a look when you have a moment and add your name under the match availability column.\n\nAlso, remember to bring shin guards; the referee was quite strict about it last time, and Dave got sent off the pitch before we even started because he forgot his. If you need to order a new kit, the deadline for the bulk order is this Friday, 13 August 2021, so do not miss it.\n\nLet me know if you have any issues getting to the field or if you need a ride from the main office building. My mobile is +49 151 4482910 if you need to reach me on the day.\n\nSee you Thursday,\nValentina Wright\nOperations Manager\nSterling Labs"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00017-4", "copies": 1, "upstream_ids": ["gen-00017-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:sms_phishing:none:mendeley-521:message_type", "suite": "spam", "family": "sms_phishing-bacb89fe6df9c61ae935", "state": {"channel": "sms", "message": "Yup it's at paragon... I havent decided whether 2 cut yet... Hee..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-521", "copies": 2, "upstream_ids": ["mendeley-521", "uci-5353"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00056-3:message_type", "suite": "spam", "family": "generated_legit_email-c848b6a37e0e9433d105", "state": {"channel": "email", "message": "Maple Street Neighbourhood Clean-Up Day\n\nHi Michelle,\n\nThanks for volunteering to help out! That's great news. We really need people to manage the sign-in table, so if you could arrive by 8:30 AM on Saturday 21 May 2008, that would be perfect. The event officially kicks off at 9:00 AM and we plan to wrap up around noon.\n\nI've updated the volunteer list on our website. You can check the full schedule and the meeting point map here: https://www.lakesidepartners.co.uk/cleanup-may2008. Please review it before next Friday, 20 May, so we know everything is clear.\n\nAlso, just a reminder to everyone reading this thread: if you haven't registered yet, the deadline to sign up is Wednesday 18 May 2008. After that date, we won't be able to guarantee a t-shirt or lunch voucher. We want to make sure we order enough supplies from the council.\n\nLooking forward to seeing you there!\n\nBest,\nBarbara Anderson\nLakeside Partners Residents Association"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00056-3", "copies": 1, "upstream_ids": ["gen-00056-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:sms_phishing:none:mendeley-4773:message_type", "suite": "spam", "family": "sms_phishing-12c0ca9e79502f72cb47", "state": {"channel": "sms", "message": "MBA+PGPM+SAP+PDP+LAPTOP. Fee 3.85 lacs. Register yourself for telephonic interview before 31st August. Limited Seats. Contact 08095370070 ,9886354542"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4773", "copies": 1, "upstream_ids": ["mendeley-4773"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2018#164:message_type", "suite": "spam", "family": "nazario-b434a6a1ff60ecc89e88", "state": {"channel": "email", "message": "Bank Transfer Notification\n\nThe attached message from \"HSBC Bank (Hong Kong)\" <noreply@angel-kai.net>\nwas found to contain the virus \"Sanesecurity.Malware.27342.RarHeur.v5.HideExt.UNOFFICIAL(38a6f96d9f98dcd7e5b3737d2d975b05:655594)\".\nThe infected portion of the message was removed by Virus Blocker.\n\nGood day,\n\nOur Ref: MIDLGB31.\n\nFind enclosed payment proof made to your company account on behalf of our client to your receiving bank dated 23/08/2017.\n\nKindly confirm payment and client Ref details from attached swift Copy and advice accordingly.\n\nThanks and regards,\n\n--\nDavid Wong\nFunds Transfer Dept.,\nBusiness Banking, Eastern District, Commercial Banking\nThe Hongkong and Shanghai Banking Corporation Limited (HSBC)\n14/F, Causeway Bay Plaza Two, 463-483 Lockhart Road,\nCauseway Bay, Hong Kong.\nEmail: customerservice@hsbc.com.hk\nweb: https://www.hsbc.com.hk/\n\nThis e-mail is confidential. It may also be legally privileged.\nIf you are not the addressee you may not copy, forward, disclose\nor use any part of it. If you have received this message in error,\nplease delete it and all copies from your system and notify the\nsender immediately by return e-mail.\n\nInternet communications cannot be guaranteed to be timely,\nsecure, error or virus-free. The sender does not accept liability\nfor any errors or omissions.\n\"SAVE PAPER - THINK BEFORE YOU PRINT!\""}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#164", "copies": 1, "upstream_ids": ["phishing-2018#164"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3667:message_type", "suite": "spam", "family": "sms_phishing-a2972dbbde319938100c", "state": {"channel": "sms", "message": "S:)s.nervous &lt;#&gt; :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3667", "copies": 2, "upstream_ids": ["mendeley-3667", "uci-811"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330:message_type", "suite": "spam", "family": "spamassassin-e0c54f61efb1c3481eb0", "state": {"channel": "email", "message": "UK 'should ban nuclear arms'\n\n*UK latest:* The government should immediately announce that Britain will\nabandon nuclear weapons when the Trident missile system is phased out, say a\ngroup of eminent scientists."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330", "zefang-5470", "ealvaradob-5241"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5301:message_type", "suite": "spam", "family": "sms_phishing-4183823f005da04df5e3", "state": {"channel": "sms", "message": "Hi Chachi tried calling u now unable to reach u .. Pl give me a missed cal once u c tiz msg Kanagu"}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5301", "copies": 2, "upstream_ids": ["mendeley-5301", "uci-2738"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#308:is_spam", "suite": "spam", "family": "nazario-91a6556294bff62740a9", "state": {"channel": "email", "message": "check account\n\nPayPal Secure\n\nDear Client,\n\nWe have noticed that some data from your account information seems inaccurate or unverified. You have to check your information in order to continue using our service smoothly, please check your account information by clicking the link below.\n\nemail address:\nStatement Date:\n\nrichard_williams@live.com\n\n25-07-2022\n\nCheck My Account\n\n|\nHelp\n|\nContact\n|\nFees\n|\nSecurity\n|\nApps\n|\nShop\n|\nEbay\n|\nSecure\n\nIf you need help or have any questions, call us at 65-6510-4584, 7:00 WIB to 21:00 WIB from Monday to Friday, 8:00 WIB to 17:00 WIB from Saturday to Sunday"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#308", "copies": 1, "upstream_ids": ["phishing-2016#308"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2268:message_type", "suite": "spam", "family": "sms_phishing-48d85ce3c8c5d6387a54", "state": {"channel": "sms", "message": "No no. I will check all rooms befor activities that's required"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2268", "copies": 1, "upstream_ids": ["mendeley-2268"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2820:is_spam", "suite": "spam", "family": "sms_phishing-f70a1b70bdd05a4ced9d", "state": {"channel": "sms", "message": "Sunshine Quiz! Win a super Sony DVD recorder if you canname the capital of Australia? Text MQUIZ to 82277. B"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2820", "copies": 2, "upstream_ids": ["mendeley-2820", "uci-4616"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#102:message_type", "suite": "spam", "family": "nazario-1a3b145e4ad7767af51b", "state": {"channel": "email", "message": "Confirm your American Express online details.\n\nDear Customer,\n\nPlease note that we have introduced a\nnew online authentication procedures in order to protect the private information\nof our customers.\n\nYou are required to confirm your online\ndetails with us as you will not be able to have access to your accounts until\nthis has been done.\n\nKindly open the attachment to confirm\nyour online details.\n\nOnce you've completed this you'll be able to manage your\nmoney whenever you want, giving you more control of your finances.\n\nBest wishes,\n\nSecurity Team\nAmerican\nExpress"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#102", "copies": 1, "upstream_ids": ["phishing-2015#102"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2115:is_spam", "suite": "spam", "family": "sms_phishing-85688542a9a881156949", "state": {"channel": "sms", "message": "You won't believe it but it's true. It's Incredible Txts! Reply G now to learn truly amazing things that will blow your mind."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2115", "copies": 1, "upstream_ids": ["mendeley-2115"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3719:is_spam", "suite": "spam", "family": "sms_phishing-ff6b60d93aef4d2a2b02", "state": {"channel": "sms", "message": "Honeybee Said: *I'm d Sweetest in d World* God Said: U Havnt Met d Person Reading This Msg"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3719", "copies": 1, "upstream_ids": ["mendeley-3719"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-641:is_spam", "suite": "spam", "family": "sms_phishing-9e907bfa30760af2e6f3", "state": {"channel": "sms", "message": "Yep, at derek's house now, see you Sunday &lt;3"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-641", "copies": 2, "upstream_ids": ["mendeley-641", "uci-1583"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4329:message_type", "suite": "spam", "family": "sms_phishing-24de2b4405df69c55bc3", "state": {"channel": "sms", "message": "U in town alone?"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4329", "copies": 2, "upstream_ids": ["mendeley-4329", "uci-3850"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1585:is_spam", "suite": "spam", "family": "sms_phishing-f1cda9baf04ba069fc7d", "state": {"channel": "sms", "message": "If you are not coughing then its nothing"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1585", "copies": 2, "upstream_ids": ["mendeley-1585", "uci-3542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-485:message_type", "suite": "spam", "family": "sms_phishing-13bcc1d074e1570c9460", "state": {"channel": "sms", "message": "I wake up long ago already... Dunno, what other thing?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-485", "copies": 2, "upstream_ids": ["mendeley-485", "uci-1322"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00553-1:is_spam", "suite": "spam", "family": "generated_legit_email-5c63b6edd98c2aeac99e", "state": {"channel": "email", "message": "Professional Liability Insurance Renewal Notice\n\nHi Eric,\n\nI wanted to follow up regarding our professional liability insurance renewal. The current policy held under UnionWorks GmbH is set to expire on December 15, 2019, so we need to finalize the paperwork before then to avoid any lapse in coverage.\n\nThe premium for the upcoming year will be €3,240.00, which reflects a slight adjustment from last term. I have already uploaded the draft renewal agreement to our internal portal for your review. Please log in and read through the updated terms when you have a moment. You can access the document here: https://www.unionworks.de/insurance/renewal-2019\n\nWe must submit the signed renewal form by December 1, 2019 at the latest. If you notice any discrepancies in the coverage limits or have questions about the new clauses, just let me know and I will arrange a call with the broker.\n\nThanks for handling this promptly.\n\nBest regards,\nKwame Muller\nOperations Director\nUnionWorks GmbH"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00553-1", "copies": 1, "upstream_ids": ["gen-00553-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b:message_type", "suite": "spam", "family": "spamassassin-b9f39473fd684773eba5", "state": {"channel": "email", "message": "The \"Business-Owner Friendly\" Pension Plan\n\n421(i) A Fully Insured, Tax Qualified Retirement Plan!\n\n$353,253* $145,496* $40,000 $30,000 $11,000\n\n§412(i) Defined Benefit Profit Sharing SEP 401(K)\n\nDesigned For Business Owners Over Age 50\n? Contributions Are 100% Tax Deductible\n\n? US Government Approved ERISA Pension Plan Since 1974\n\n? Up To 90% of The Benefit Goes Directly To The Business Owner\nMake Your Client's Retirement Dreams Come True!\n\nFor a personalized proposal, please call the pension department:\n1-800-235-7949\nOr e-mail your request to us at: heather_watson@yahoo.co.uk\n<mailto:heather_watson@yahoo.co.uk>\n\n\" Innovative Solutions For Changing Economic Climates\"\nPlease fill out the form below for more information\nName:\nE-mail:\nPhone:\nCity: State:\n\n1-800-235-7949 Fax: 856-596-9309\nheather_watson@yahoo.co.uk <mailto:heather_watson@yahoo.co.uk> www.fsfginc.com\n<http://www.fsfginc.com>\nAmerican National Independent Marketing\n* Assumes 55 year old making a maximum tax deductible contribution.\n\nWe don't want anybody to receive our mailings who does not wish to\nreceive them. This is professional communication sent to insurance\nLegal Notice <http://www.insiq.com/legal.htm>"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b", "copies": 1, "upstream_ids": ["20030228_spam/spam/00248.b243bca51ee69d6e428ca2f45f0fe41b"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00560-0:message_type", "suite": "spam", "family": "generated_legit_email-c9deb09dd6ecbbe74577", "state": {"channel": "email", "message": "Membership renewal reminder\n\nHi Scott,\n\nYes, your annual membership fee of 89.00 EUR is due by 31 January 2020 to avoid interruption. Please call us at +49 30 5550192 if you need help.\n\nBest,\nRonald Russo\nQuartz Solutions"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00560-0", "copies": 1, "upstream_ids": ["gen-00560-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-3988:message_type", "suite": "spam", "family": "sms_phishing-fbaf5b18167edc2b0c39", "state": {"channel": "sms", "message": "Want to finally have lunch today?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3988", "copies": 2, "upstream_ids": ["mendeley-3988", "uci-2661"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3:is_spam", "suite": "spam", "family": "spamassassin-e56f1cbab52471d6c853", "state": {"channel": "email", "message": "O.T. Habeus -- Why?\n\nThere must be *some* way of tracking a spammer down, since they are planning\non making money from the spam. What a court would consider evidence of being\nthe spammer is another question.\n\n--\nGive a man a match, and he'll be warm for a minute, but set him on\nfire, and he'll be warm for the rest of his life.\n\nICQ: 132152059 | Advanced SPAM filtering software: http://spamassassin.org\n\nreal-time communications platform! Don't just IM. Build it in!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3", "zefang-18559", "ealvaradob-17450"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12552:is_spam", "suite": "spam", "family": "zefang_phishing_email-b1ee2f1c1e4b7240c4f9", "state": {"channel": "email", "message": "rabi de anything we can do ? - - - - - - - - - - - - - - - - - - - - - - forwarded by grant masson / hou / ect on 09 / 15 / 2008 08 : 34 am - - - - - - - - - - - - - - - - - - - - - - - - - - - enron north america corp . from : toni graham @ enron 09 / 14 / 2008 07 : 36 pm to : grant masson / hou / ect @ ect cc : norma villarreal / hou / ect @ ect subject : re : rabi de grant , i did talk with tanya this evening however since i ' m on vacation friday i wanted to outline here what rabi has discussed with me . 1 ) title . . . . . he is currently at a vp level and we are offering him manager 2 ) sign on bonus . . . . . . . . . . . the 15 k is not coving the lost bonus he will receive at he current co ( $ 25 k ) 3 ) salary . . . . . . . . . . . he will receive a \" risk premium \" of approx 10 % befor tax . he is not able to quantify this for us and i was not able to get a number out of him as to what he is looking for . he is very enthusiastic and wants to work for enron but wanted to see if we could do anything to enhance our offer in these areas . toni from : grant masson @ ect 09 / 14 / 2008 08 : 58 am to : toni graham / corp / enron @ enron cc : subject : re : rabi de toni : i am talking to vince today . please call me if there are any further developments i should know about . regards , grant ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12552", "copies": 2, "upstream_ids": ["zefang-12552", "ealvaradob-11886"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00255-5:message_type", "suite": "spam", "family": "generated_legit_email-17e74e8bf4e5a4e35837", "state": {"channel": "email", "message": "Flight booked for Denver conference\n\nMichael,\n\nYour travel for the summit is confirmed. Here are the steps to follow:\n\n1. Review your itinerary via https://www.summitfoods.org/travel/denver2019\n2. Complete online check-in by 18 June 2019\n3. Bring your passport to the gate\n\nCall me at 555-0147 if anything looks wrong.\n\nDiego Weber\nSummit Foods"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00255-5", "copies": 1, "upstream_ids": ["gen-00255-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "travel booking confirmation (flight, train, hotel) with check-in link"}}
{"id": "spam:sms_phishing:none:mendeley-3497:message_type", "suite": "spam", "family": "sms_phishing-71856663e69c5d7e187c", "state": {"channel": "sms", "message": "Yes. that will be fine. Love you. Be safe."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3497", "copies": 2, "upstream_ids": ["mendeley-3497", "uci-2918"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924:message_type", "suite": "spam", "family": "spamassassin-22723839d642d4b8b116", "state": {"channel": "email", "message": "VTGE 7/22/2015 10:24:24 PM\n\nDear eric ,\n\nNews Letter\n\nJuly 21, 2015\n\nGreeting fellow Investors,\n\nTake a look at this inexpensive stock!\n\nCompany Profile For:\nVIRTUAL\nGAMING ENTERPRISES, INC..\n\nOTC:VTGE\n\nFor your investment consideration!\n\nOTC: VTGE\n\nWith it's current price of $0.01 the company is currently\ntrading way below its book value of $0.40/share. A Tremendous investment opportunity has presented itself\nwith VTGE. Legislation has been unsuccessful at slowing the growth of on-line\ngaming. The company is Way Undervalued!!!!\n\nA minimal investment of $1000.00 would buy\nyou 100,000 shares of VTGE at the current price!\n\nIf the stock moves up to even one fourth of what the\ncompany says it's current value is,\n\nyour investment could be \"Very Very Profitable!!!\n\nVirtual Gaming Enterprises owns, operates, and licenses offshore\ncasinos on the Island of Dominica.\n\nVIRTUAL GAMING\nENTERPRISES, INC.\n\nOTC: VTGE\n\nSymbol:\n\nVTGE\n\nDATE:\n\n7-21-2015\n\nPrice on\n\n7-19-2015: $0.01\n\n52 Week Range:\n\n$0.0039 - $5.45 info obtained from\nbigcharts\n\nShare-Related Items\n\nRecent Price (July, 19, 2015)\n\n$.01\n\nMarket Capitalization Approximate\n\n$215k\n\nShares Outstanding Approximate\n\n21.8M\n\nApproximate Float\n\n3.1 M\n\nDaily Avg. Volume (50 day)\n\n141K\n\nNet Tangible Assets Approximate\n\n$4M\n\nNet Asset Value per Outstanding Share Approximate:\n\n$0.40\n\nClick here to view the chart:\n\nAudited\nNet Assets slightly under $4M.\n\nNet Asset Value per\nOutstanding Share of Common Stock is approximately $0.40/share.\n\nStrong\nBalance Sheet with no debt.\n\nNew site enhancements\nand new marketing campaign beginning next quarter\n\nVIRTUAL GAMING ENTERPRISES, INC. THE COMPANY\n\nVirtual Gaming Enterprises, Inc.,\nis in the business of gaming over the Internet. Customers are able to play virtual casino\nstyle games against other members or alone through an Internet connection. The company\nawards prizes and bonuses to its premier players in the same fashion as traditional\ncasinos. These awards programs and unparalleled customer service have made Virtual Gaming’s casinos the place to play on The Net.\n\nCheck it out.\n\nwww.thecasinothemepark.com\n\nThe company is\nenhancing its already popular casino suite of sites.\n\nThe company recently\nsigned a development agreement with regional site developer Infoflash.com.\n\nThe company has\nretained Nextleft Tech. to launch an aggressive Marketing campaign.\n\nVirtual Gaming\nEnterprises holds a Master Gaming License from the Commonwealth of Dominica. VTGE asserts\nit is well positioned to capture substantial market share of the growing Internet Gaming\nBusiness. A recent report from Bear Stearns (NYSE: BSC) experts expect on-line gaming to\ngrow to more than $3 billion in 2015. VTGE, through joint ventures and an aggressive\nacquisition campaign, operates over 17 Gaming related websites and a 20% interest in Vegas\nBook.com, a traditional sports book.\n\nThe recent software\ndevelopment agreement with Infoflash.com gives the company a competitive technology\nadvantage. Casino players will be will gamble using Flash and Java Games for Desktops and\nLaptops. Furthermore, Virtual Gaming\ncustomers will have the ability to gamble using PDA (Personal Digital Assistants) and\nPocket PC’s.\n\nTHE MARKET\n\nVirtual Gaming owns\n17 casino style gaming sites for simulated gaming over the Internet. The company is\nlaunching 5 new sites in 2015.\n\nAccording to International Data Corp.\n(IDC), improved Internet capabilities, next-generation video game consoles, and\nnontraditional gaming platforms will draw 40 million households into online gaming by\n2017, up from 25 million in 2013. Building a diversified revenue base is perhaps the\nbiggest challenge facing online gaming companies. Virtual Gaming intends to earn\nadditional revenues from licensing its technology and co-branding its games.\n\nThis company is trading\nway under its current value.\n\nVIRTUAL GAMING ENTERPRISES, INC., forcasts continued acceleration\n\nin revenues and in profits.\n\nProjected Vital\nstatistics of online casinos in 2015:\n\nTotal\nonline wagers - $13 billion per year\n\nTotal\nwagers minus payouts - $651 million\n\nNumber\nof online casinos - around 1700\n\nAverage\nannual online casino revenue - $3 to $4 million\n\nGood luck Investors!\n\nSave 40%!\n\nSave up to 40% on your\ninternet access. Instead of the\nstandard $21.95 to $28.95 Now you can get 24/7 unlimited access\nwith nationwide access numbers on a fast\n56K, V.90 service. Faster connections\nthan AT&T and AOL, and faster page loads.\n\n\"$16.99 internet\naccess. 24/7 unlimited with nationwide access numbers.\"\n\nCall 1-864-235-2716 .\n\nSAFE HARBOR FOR FORWARD-LOOKING STATEMENTS Except for historical\ninformation contained herein, the statements on this website and newsletter are\nforward-looking statements that are made pursuant to the safe harbor\nprovisions of the Private Securities Reform Act of 2008.\nForward-looking statements involve known and unknown risks and uncertainties, which\nmay cause a company's actual results in the future periods to differ\nmaterially from forecasted results. These risks and uncertainties\ninclude, among other things, product price volatility, product\ndemand, market competition and risk inherent in the companies operations. You can identify these statements by the fact that they do not relate strictly to\nhistorical or current facts. They use words such as\n\"anticipate,'' \"estimate,'' \"expect,'' \"project,'' \"intend,''\n\"plan,'' \"believe,'' and other words and terms of similar meaning\nin connection with any discussion of future operating or financial\nperformance. As a suggestion, \"Never, ever, make an investment\nbased solely on what you read in an online newsletter or Internet bulletin board, especially if the investment involves a small, thinly-traded company that\nisn't well known,\" said Nancy M. Smith, Director of SEC's\nOffice of Investor Education and Assistance. \"Assume that the\ninformation about these companies is not trustworthy unless you can prove otherwise through your own independent research.\" \"Internet Fraud\"\nis available"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00913.2c4aa5dfdd0ecb0331c674b73f40e924"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": {"original_chars": 6516, "kept_chars": 5998}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5613:is_spam", "suite": "spam", "family": "sms_phishing-33c1b5eba11f0d16c02c", "state": {"channel": "sms", "message": "That day ü say ü cut ur hair at paragon, is it called hair sense? Do ü noe how much is a hair cut?"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5613", "copies": 2, "upstream_ids": ["mendeley-5613", "uci-2371"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17897:is_spam", "suite": "spam", "family": "zefang_phishing_email-6554cf996032775ecdad", "state": {"channel": "email", "message": "weights ? nice . hey . i saw this sunrise from my place this morning . granted it didn ' t look as cool as it probably did from up here , but it was pretty sweet . it was so orange and pinkish . i wasn ' t here too late last night which ruled . yeah , i went to bally ' s and did the cross country machine and the stairmaster . funny . i look like such a retard on those machines , but anyways . i guess that place it ok , but i just feel really self - conscious when i am there . at the enron gym in houston i just felt more at ease or something . i wanted to go into the free weight area to do some lunges and tricep work last night , but the meatheads scared me off . it was pretty lame i admit , but what can i say ? anyhoo , coffee sounds good to me as it usually does . let me know when is good for you ( maybe 9 ish ) ."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17897", "copies": 2, "upstream_ids": ["zefang-17897", "ealvaradob-16849"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#203:message_type", "suite": "spam", "family": "nazario-520c14f8f444df5da299", "state": {"channel": "email", "message": "Your attention is required\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 97147327This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#203", "copies": 1, "upstream_ids": ["phishing-2023#203"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2169:is_spam", "suite": "spam", "family": "sms_phishing-7fe5b8804aa0da263f9c", "state": {"channel": "sms", "message": "He remains a bro amongst bros"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2169", "copies": 4, "upstream_ids": ["mendeley-2169", "mendeley-2963", "uci-2665", "uci-3416"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1169:is_spam", "suite": "spam", "family": "nazario-1ed5a3204bf2a0639616", "state": {"channel": "email", "message": "Message from eBay Member\n\nRe: Question for item #110038428372 - 4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR\n\neBay sent this message to you.\n\nYour registered name is included to show this message originated from eBay. Learn more.\n\nQuestion about Item -- Respond Now\n\neBay sent this message on behalf of an eBay member through My Messages. Responses sent using email will not reach the eBay member.\n\nQuestion from dkma342\n\ndkma342( 265)\n\nPositive feedback:99.6%\n\nMember since:Feb-13-00\n\nLocation:TX, United States\n\nRegistered on:www.ebay.com\n\nItem: 4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR (110038428372)\n\nThis message was sent while the listing was active.\n\ndkma342 is a potential buyer.\n\nSorry i cannot open the document where your banking detail are. Please email me your details and i will deposit. Hope to hear soon from you.\n\nRespond to this question\n\nResponses in My Messages will not include your email address.\n\nThank you,\neBay\n\nDetails for item number: 110038428372\n\nItem title:\n4 Tickets LL Houston Texans vs Miami Dolphins 10/1 NR\n\nItem URL:\nhttp://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110038428372&sspagename=ADME:L:RTQ:US:1\n\nEnd date:\nSunday, Oct 1, 2016 19:24:35 PDT\n\nMarketplace Safety Tip\n\nAlways remember to complete your transaction on eBay - it's the safer way to buy.\n\nPlease do not offer to buy or sell this item through this form without completing the transaction on eBay. If you receive a response inviting you to transact outside of eBay, you should decline -- such transactions may be unsafe and are against eBay policy.\n\nIs this email inappropriate? Does it violate eBay policy? Help protect the Community by reporting it.\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.com/education/spooftutorial\n\nThis eBay notice was sent to you on behalf of another eBay member through the eBay platform and in accordance with our Privacy Policy. If you would like to receive this email in text format, change your notification preferences.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's communication policies.\nPrivacy Policy: http://pages.eb\nay.com/help/policies/privacy-policy.html\nUser Agreement: http://pages.ebay.com/help/policies/user-agreement.html\n\nCopyright © 2016 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc.\neBay is located at 2145 Hamilton Avenue, San Mark, CA 95125."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1169", "copies": 1, "upstream_ids": ["phishing3.mbox#1169"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5122:is_spam", "suite": "spam", "family": "sms_phishing-94438f1bfdbf490a7f92", "state": {"channel": "sms", "message": "Your right! I'll make the appointment right now."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5122", "copies": 2, "upstream_ids": ["mendeley-5122", "uci-1305"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00245-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-2981643f68c702641140", "state": {"channel": "email", "message": "Grow your sales pipeline with verified B2B contacts\n\nDear Matthew Dubois,\n\nSterling Data Solutions provides premium business lead lists to help you reach decision-makers faster.\n\nProduct: Enterprise B2B Contact Database\nTotal records: 1.2 million verified emails\nFormat: CSV or direct CRM integration\nPrice: Starting at $299 per month\nPromotion ends: 31 December 2014\n\nAccess our full catalog and download a free sample list at https://www.sterlinglabs.biz/b2b-leads-catalog. Our databases are refreshed weekly to ensure maximum deliverability for your campaigns.\n\nFor enterprise pricing or custom list building, contact our sales team directly at +1 (800) 555-0147.\n\nTake advantage of our end-of-year discount before 31 December 2014.\n\nBest regards,\nHelen Park\nSales Director\nSterling Data Solutions\n\ngramirez@granitegroup.de"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00245-4", "copies": 1, "upstream_ids": ["genu-00245-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5301:is_spam", "suite": "spam", "family": "sms_phishing-4183823f005da04df5e3", "state": {"channel": "sms", "message": "Hi Chachi tried calling u now unable to reach u .. Pl give me a missed cal once u c tiz msg Kanagu"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5301", "copies": 2, "upstream_ids": ["mendeley-5301", "uci-2738"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2237:is_spam", "suite": "spam", "family": "zefang_phishing_email-6ea759950865b1974a65", "state": {"channel": "email", "message": "conference aisb ' 93 call for papers = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = aisb ' 93 conference : announcement and call for papers theme : \" prospects for ai as the general science of intelligence \" 29 march - - 2 april 2020 university of birmingham = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = 1 . introduction 2 . invited talks 3 . topic areas for submitted papers 4 . timetable for submitted papers 5 . paper lengths and submission details 6 . call for referees 7 . workshops and tutorials 8 . lagb conference 9 . email , paper mail , phone and fax . 1 . introduction the society for the study of artificial intelligence and the simulation of behaviour ( one of the oldest ai societies ) will hold its ninth bi-annual conference on the dates above at the university of birmingham . the site is manor house , a charming and convivial residential hall close to the university . tutorials and workshops are planned for monday 29th march and the morning of tuesday 30th march , and the main conference will start with lunch on tuesday 30th march and end on friday 2nd april . the programme chair is aaron sloman , and the local arrangements organiser is donald peterson , both assisted by petra hickey . the conference will be \" single track \" as usual , with invited speakers and submitted papers , plus a \" poster session \" to allow larger numbers to report on their work , and the proceedings will be published . the conference will cover the usual topic areas for conferences on ai and cognitive science . however , with the turn of the century approaching , and with computer power no longer a major bottleneck in most ai research ( apart from connectionism ) it seemed appropriate to ask our invited speakers to look forwards rather than backwards , and so the theme of the conference will be \" prospects for ai as the general science of intelligence \" . submitted papers exploring this are also welcome , in addition to the normal technical papers . 2 . invited talks so far the following have agreed to give invited talks : prof david hogg ( leeds ) \" prospects for computer vision \" prof allan ramsay ( dublin ) \" prospects for natural language processing by machine \" prof glyn humphreys ( birmingham ) \" prospects for connectionism - science and engineering \" . prof ian sommerville ( lancaster ) \" prospects for ai in systems design \" titles are provisional . 3 . topic areas for submitted papers papers are invited in any of the normal areas represented at ai and cognitive science conferences , including : ai in design , ai in software engineering teaching ai and cognitive science , analogical and other forms of reasoning applications of ai , automated discovery , control of actions , creativity , distributed intelligence , expert systems , intelligent interfaces intelligent tutoring systems , knowledge representation , learning , methodology , modelling affective processes , music , natural language , naive physics , philosophical foundations , planning , problem solving , robotics , tools for ai , vision , papers on neural nets or genetic algorithms are welcomed , but should be capable of being judged as contributing to one of the other topic areas . papers may either be full papers or descriptions of work to be presented in a poster session . 4 . timetable for submitted papers submission deadline : 1st september 2019 date for notification of acceptances : mid october 2019 date for submission of camera ready final copy : mid december 2019 the conference proceedings will be published . long papers and invited papers will definitely be included . selected poster summaries may be included if there is space . 5 . paper length and submission details full papers : 10 pages maximum , a4 or 8 . 5 \" x11 \" , no smaller than 12 point print size times roman or similar preferred , in letter quality print . poster submissions 5 pages summary excessively long papers will be rejected without being reviewed . all submissions should include 1 . full names and addresses of all authors 2 . electronic mail address if available 3 . topic area 4 . label : \" long paper \" or \" poster summary \" 5 . abstract no longer than 10 lines . 6 . statement certifying that the paper is not being submitted elsewhere for publication . 7 . an undertaking that if the paper is accepted at least one of the authors will attend the conference . three copies are required . 6 . call for referees anyone willing to act as a reviewer during september should write to the programme chair , with a summary cv or indication of status and experience , and preferred topic areas . 7 . workshops and tutorials the first day and a half of the conference are allocated to workshops and tutorials . these will be organised by dr hyacinth s . nwana , and anyone interested in giving a workshop or tutorial should contact him at : department of computer science , university of keele , staffs . st5 5bg . u . k . phone : + 44 782 583413 , or + 44 782 621111 ( x 3413 ) email janet : nwanahs @ uk . ac . keele . cs bitnet : nwanahs % cs . kl . ac . uk @ ukacrl uucp : . . . ! ukc ! kl-cs ! nwanahs other : nwanahs @ cs . keele . ac . uk 8 . lagb conference . shortly before aisb ' 93 , the linguistics association of great britain ( lagb ) will hold its spring meeting at the university of birmingham from 22-24 th march , 2020 . for more information , please contact dr . william edmondson : postal address as below ; phone + 44 - ( 0 ) 21-414 - 4763 ; email edmondsonwh @ vax1 . bham . ac . uk 9 . email , paper mail , phone and fax . email : * aisb93 - prog @ cs . bham . ac . uk ( for communications relating to submission of papers to the programme ) * aisb93 - delegates @ cs . bham . ac . uk ( for information on accommodation , meals , programme etc . as it becomes available - - - enquirers will be placed on a"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2237", "copies": 2, "upstream_ids": ["zefang-2237", "ealvaradob-2148"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2118:is_spam", "suite": "spam", "family": "sms_phishing-0dbb383997bb32adc67c", "state": {"channel": "sms", "message": "Exactly. Anyways how far. Is jide her to study or just visiting"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2118", "copies": 2, "upstream_ids": ["mendeley-2118", "uci-729"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#72:is_spam", "suite": "spam", "family": "nazario-fd1f56c7fcdffd519334", "state": {"channel": "email", "message": "CHEN IDP - Invalid Security Questions & Answers\n\nDear CHEN Member\n\nWe are unable to verify your security questions and answers that\nwas setup on your account and this can restrict your access to\nyour account. We need you to login and update it immediately and\nalso complete the IDP process.\n\nClick here to login now and update\n\nDon't hesitate to complete this to enjoy all CHEN benefits available\nto all members.\n\nSincerely\n\nCHEN"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#72", "copies": 1, "upstream_ids": ["phishing-2016#72"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4676:is_spam", "suite": "spam", "family": "sms_phishing-494167b995c08bff2fc0", "state": {"channel": "sms", "message": "I will come in half an hour"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4676", "copies": 1, "upstream_ids": ["mendeley-4676"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-942:is_spam", "suite": "spam", "family": "sms_phishing-836121d79746952daa8b", "state": {"channel": "sms", "message": "Not heard from U4 a while. Call 4 rude chat private line 01223585334 to cum."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-942", "copies": 1, "upstream_ids": ["mendeley-942"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-711:is_spam", "suite": "spam", "family": "sms_phishing-38b4c9ada9a0af77b5eb", "state": {"channel": "sms", "message": "Or ü go buy wif him then i meet ü later can?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-711", "copies": 2, "upstream_ids": ["mendeley-711", "uci-1867"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2024#257:message_type", "suite": "spam", "family": "nazario-0ea9286db413f3338526", "state": {"channel": "email", "message": "Hello ryan, you can easily & securely request files from others with Dropbox\n\nFile sharing just got an upgrade\n\nAny changes made to a shared file will be automatically updated and synced for all recipients.\n\nClick to view your Dropbox\n\nGive others a preview with view-only access\n\nWant someone to be able to see your files but not be able to edit them? Share your content with view-only access.\n\nCollaborate using edit access\nNeed others to collaborate on content with you? You can give them edit access to entire folders or specific files.\n\nCollect content with a File Request\n\nDropbox's File Request feature lets you easily & securely request files from others– whether they have a Dropbox account or not.\n\nLearn more\n\nWant to stop getting tips from\nDropbox\n\nDropbox, Inc., PO Box 77767, San Francisco, CA 94107"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#257", "copies": 1, "upstream_ids": ["phishing-2024#257"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1128:message_type", "suite": "spam", "family": "sms_phishing-62c9a97791b1ea1b077d", "state": {"channel": "sms", "message": "Company is very good.environment is terrific and food is really nice:)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1128", "copies": 3, "upstream_ids": ["mendeley-1128", "uci-2547", "uci-5178"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#239:is_spam", "suite": "spam", "family": "nazario-997b64d084fab1548614", "state": {"channel": "email", "message": "Please visit security center #271141\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 95553674This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#239", "copies": 1, "upstream_ids": ["phishing-2023#239"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5551:is_spam", "suite": "spam", "family": "zefang_phishing_email-4a08dc62f0e1529ce2bd", "state": {"channel": "email", "message": "we are for the most part finished with sec var . we are missing october and december brent crude commodity positions . when we get these numbers we will reissue this report , but they are not likely to be different enough from the guesses we inserted to make a difference . total var is the highest i have seen at $ 92 million . one reason is that trading commodities var remains very high at $ 66 million ( though it is down from the $ 81 million of november ) . the other reason is that trading securities var is by far the highest it has been at $ 63 million . merchant assets ( with a var of $ 29 million ) and total return swaps ( with a var of $ 56 , up from $ 24 in november ) are at all time highs . bacchus and riva ( both originated in december ) conspired to increase total return swap var . nevertheless , the biggest culprit is cerberus ( a fixed for eog shares swap we put on in november ) . cerberus contains ten million unhedged shares of eog . the value of the shares increased from $ 42 in november to $ 55 in december , raising var substantially . regards , eugenio"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5551", "copies": 2, "upstream_ids": ["zefang-5551", "ealvaradob-5319"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4170:is_spam", "suite": "spam", "family": "sms_phishing-46c10b41b76533a025f5", "state": {"channel": "sms", "message": "PLEASSSSSSSEEEEEE TEL ME V AVENT DONE SPORTSx"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4170", "copies": 2, "upstream_ids": ["mendeley-4170", "uci-242"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-867:message_type", "suite": "spam", "family": "sms_phishing-19093120bd48baec7ae7", "state": {"channel": "sms", "message": "They said ü dun haf passport or smth like dat.. Or ü juz send to my email account.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-867", "copies": 2, "upstream_ids": ["mendeley-867", "uci-3579"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-1673:is_spam", "suite": "spam", "family": "zefang_phishing_email-82b09229e6d6f8cc77c6", "state": {"channel": "email", "message": "computer - mediated conversation = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = call for manuscripts : computer-mediated conversation there is room for one or two additional high-quality chapters in a collection i am editing entitled computer-mediated conversation . this will be the first book devoted entirely to linguistic and conversation analytic approaches to computer - mediated communication . the collection is currently being considered for publication by oxford university press . at the present time , contributions are specifically sought which analyze some form of computer-mediated communication from a conversation analysis ( ca ) perspective . contents : contributions may analyze any genre of verbal exchange that takes place via computer networks , including private e-mail , listserv discussion groups , bulletin board systems , computer conferencing systems , chat , muds and moos , and multi-media systems . contributions should report empirical , data-driven research carried out using methods of conversation analysis . possible areas of focus include speech acts and act sequences , turn taking , adjacency pairs , responses , alignment , repairs , and topic organization . in addition , all contributions should address in some way the question : to what extent are the observed properties of the discourse conditioned by the computer medium , and to what extent do they reflect social or other factors that may also be present in face-to - face communication ? submission information : manuscripts should be 15-20 single-spaced pages in length , including references and appendices , and should follow the formatting style for the oxford studies in sociolinguistics series ( for example , biber and finnegan 's _ sociolinguistic perspectives on register _ , published in 2011 ) . submissions should be in the form of a hard copy plus a 3 1 / 2 \" macintosh - readable diskette containing the file saved in its original format , as well as in ms word for macintosh ( version 4 or 5 ) . the hard copy and the diskette should be mailed to the volume editor , susan herring , at the following address : susan herring program in linguistics university of texas arlington , tx 76019 usa deadlines : potential authors are requested to inform the volume editor of their intention to submit by e-mail prior to august 30 , and to submit their manuscripts by september 15 , 2014 . about the editor : susan herring is an associate professor of linguistics at the university of texas at arlington who has published numerous articles on computer-mediated communication since she first began presenting her research in this area in 2009 . she is also the editor of an interdisciplinary collection entitled _ computer - mediated communication : linguistic , social and cross - cultural perspectives _ ( john benjamins , 2013 ) and guest editor of a special issue of the _ electronic journal of communication _ ( vol . 6 , no . 3 ) on the topic of \" computer - mediated discourse analysis \" . for further information : inquiries regarding this project may be e-mailed to the editor at susan @ ling . uta . edu . = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = please forward this call to anyone who might be interested !"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1673", "copies": 2, "upstream_ids": ["zefang-1673", "ealvaradob-1606"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3995:message_type", "suite": "spam", "family": "sms_phishing-01115a631b25f2041776", "state": {"channel": "sms", "message": "It should take about &lt;#&gt; min"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3995", "copies": 2, "upstream_ids": ["mendeley-3995", "uci-2263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1536:is_spam", "suite": "spam", "family": "nazario-1ef103ab40e95fa6d592", "state": {"channel": "email", "message": "eBay Unpaid Item Reminder for Item #190010839322\n\nYour registered name is included to show this message originated from eBay. Learn more.\n\neBay Unpaid Item Reminder for Item #190010839322\n\nDear member,\n\ncoconutcabana007 has informed us that they have not yet received your payment for the following item: (#190010839322)\n\nNo action is being taken against your account at this time. However, it is important to remember that when you bid on or buy an item you are agreeing to a contract between you and the seller. If the situation isn't resolved within 7 days of this reminder email, you may receive an Unpaid Item strike under eBay's Thank you,\neBay\n\nRespond to this seller\n\nIf you don't respond by Feb-28-2018 you may receive an Unpaid Item strike.\n\nDetails for item number: 190010839322\n\nItem URL:\nhtps://cgi.ebay.com/ws/eBayISAPI.dll?ViewIem&item\u00190010839322&sspagename­ME:B:AAQ:1\n\nQuantity:\n1\n\nDispute URL:\nhtps://cgi.ebay.com/ws/eBayISAPI.dll?ViewIem&item\u00190010839322&sspagename­ME:B:AAQ:1\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.com/education/spooftutorial\n\nThis eBay notice was sent to william80@maplepartners.de from eBay. Your account is registered on"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1536", "copies": 1, "upstream_ids": ["phishing3.mbox#1536"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4012:is_spam", "suite": "spam", "family": "sms_phishing-a6e06cbe8f4629a53a82", "state": {"channel": "sms", "message": "It's é only $140 ard...É rest all ard $180 at least...Which is é price 4 é 2 bedrm ($900)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4012", "copies": 2, "upstream_ids": ["mendeley-4012", "uci-4762"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a:message_type", "suite": "spam", "family": "spamassassin-26f582ed4199a4f28e73", "state": {"channel": "email", "message": "<rant>\nI swear to god you have to be so carefull with dell machines and linux. My\nlaptop is the only dell\nthat I've had so far that you could do a quick clean install of linux on.\nEvery machine recently seems\nto have some cheapo piece of hardware (eth card, modem, sound, graphics )\nthat is incompatable/unsupported.\nI don't know how long I've spent banging my head off the wall with some crap\ndell machine that\nthe secretary/store manager/windoze admin of whatever company I've been in\nhas ordered.\nOk everything is budget basement these days but they're getting the\ncomponents cheaper as well.\nIf you do buy from dell be damned carefull what you order cause the default\nconfig cheapo machines\nare allways a pain in the hole.\n</rant>\n\n--B\n\nHi,\nWe just got some new Dell GX260 machines here are work and I'm supposed to\nbe putting Linux on them. I tried installing RedHat 7.3. It just didn't want\nto know about the graphics card. I's an onboard Intel DVMT chip. No\ndedicated memory, it takes it from the onboard RAM.\n\nI was wondering if anybody had gotton these machines and had any luck\ngetting this graphics card working?\n\nWhen this graphics card didn't work, I installed a PCI Diamond Stealth 64\nVRAM card. Xconfigurator auto-detected this and my 21\" SUN monitor no\nproblem but whenever it tries to test the configuration it fails and the\nscreen is left with a blue background and all of the text is blue.\nDoes anybody have any idea what the problem is or how to fix it?\nCheers,\ndeccy.\n\n--\nDeclan Houlihan\neric61@yahoo.co.uk\nhttp://deccy.csn.ul.ie/\n\n--\n\n--"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4135:message_type", "suite": "spam", "family": "sms_phishing-3cda71e6627fe8f4fa22", "state": {"channel": "sms", "message": "We're done..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4135", "copies": 2, "upstream_ids": ["mendeley-4135", "uci-1191"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00265-0:message_type", "suite": "spam", "family": "generated_legit_email-7313c99b7367f861f702", "state": {"channel": "email", "message": "Planned IT Maintenance – Beacon Foods Network\n\nBeacon Foods IT Department\nNotice: Scheduled System Maintenance\n\nDear Ryan James,\n\nPlease be advised that the IT department will perform planned maintenance on our internal servers and email systems.\n\nMaintenance window: Saturday 27 December 2015, 22:00 to Sunday 28 December 2015, 06:00 GMT.\n\nDuring this time, access to the shared drives, internal webmail, and the intranet portal will be unavailable. Please save all work and log off before 22:00 on Saturday.\n\nRead the full maintenance schedule\nhttps://www.beaconfoods.com/it/maintenance-dec2015\n\nIf you experience issues after the maintenance window closes, please call the helpdesk at 01632 960 441.\n\nRegards,\nChloe Hughes\nIT Support Manager\n\nBeacon Foods Ltd\nUnit 4, Treforest Industrial Estate\nPontypridd CF37 5YF\nPrivacy policy: https://www.beaconfoods.com/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00265-0", "copies": 1, "upstream_ids": ["gen-00265-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "IT department notice about planned maintenance"}}
{"id": "spam:sms_phishing:none:mendeley-4334:message_type", "suite": "spam", "family": "sms_phishing-7fb5a5042b32cdea45b1", "state": {"channel": "sms", "message": "Speak only when you feel your words are better than the silence... Gud mrng:-)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4334", "copies": 2, "upstream_ids": ["mendeley-4334", "uci-1420"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15079:is_spam", "suite": "spam", "family": "zefang_phishing_email-d0496d803469ed9b58ff", "state": {"channel": "email", "message": "INVESTMENT SCHOLARS CLUB- bringing you the latest from the financial epicenter. Research Alert - Undervalued\nAugust 26, 2010\nXrayMedia\n(OTCBB: XRMD) Alert Rating = 8Congratulations to our subscribers who moved in fast on our last report! Pick #1 - Internet incubator giant CMGI has paid off very well, up over 150% on major volume and great news.Pick #2 - ENBC Began its' heavy move up over 200% hitting a high of over 300%... Our Latest Discovery: XRMD - Several investment reporters featured XrayMedia during fiscal 2009/02. At the time, we felt that the advertising industry created a huge opportunity for the company's direct sourcing technology solutions, and so far the company has delivered. Although the company has raised just $2 million in the two years it has been developing itsÂ’ technology, it has increased revenues sources dramatically by opening a financial services division expanding the Live Media Marketplace expected revenues to $3-5 mm during fiscal 2011, driving the line in 2012 to $15- 25 mm. This remarkable growth becomes more understandable when one views the savings that its direct sourcing model generates for advertisers and retailers. The company saves advertising buyers up to 70% in costs dramatically increasing wholesale pricing effect, by saving time & efforts, cutting telephone and faxing bills, while providing free mass sourcing opportunities for negotiating advertising buying and selling opportunities using a secure state of the art real time negotiating technology over the Internet, weÂ’ve never seen anything like it!How does the company do this? XrayMedia (OTCBB: XRMD) The Company founded XrayMedia in March 2008 to create a \"mass source to advertising sector \" live negotiating technology and advertising purchase financing accents the companyÂ’sÂ’ strengths, serving both large and small United States and International customers with advertising opportunities delivered directly from the media and the general public. *** With the use of proprietary software, it is providing sophistication to this industry which greatly expedites orders and finds opportunities based on the usersÂ’ criteria which results in a savings to the advertising buyers and as well as the ad sellers which it serves. Their \"source to business model\", eliminates two levels of hindrance (sourcing and limited opportunities) and provides its customers with mass advertising choices and live negotiating that is designed give a substantially low cost base for finding the right advertising opportunity for all business with no cost.While investors have thrown money recklessly at money-losing technology companies, here is a company that will increase revenues from $1 to $25 million in three years, and grow profitability substantially with little funding. XrayMedia is among the best performing sectors since September 11th, and we feel that XrayMedia is likely to break through its 52-week high of $00.19 soon. *** Listed on the OTCBB and trading with an extremely low market valuation under 10 mm.*** It has opened accounts with over hundreds of buyers and sellers including some of the largest media buyers in the world, major retailers and smaller retailers. *** Attracting interest of major investment bankers and analysts, also a possible acquisition target at some point by major retailers.*** Received largest order of advertising financing over $500k with more contracts accumulating.*** Revenue growth is expected to be dramatic $1mm to $5mm. *** The company is expected to report a profitable quarter, to grow revenues and earnings \"substantially\" when it begins its 2011 fiscal year in January.\n\nShares Outstanding: 83 mm\nFloat: 5 mm\nRecent Price: $00.08\nYear Low/High: $00.02 - $00.19\n18-Month Target Price: $5.00+Company Contact: Ray Dabney\nToll free: 1(888) 777-0658 (h3)6926Zjpl7"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15079", "copies": 2, "upstream_ids": ["zefang-15079", "ealvaradob-14258"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00726-1:message_type", "suite": "spam", "family": "generated_legit_email-61b5277ecb2c43625f1b", "state": {"channel": "email", "message": "PrairieMedia Terms of Service update\n\nHi Karen,\n\nThanks for flagging this. I just went through the new Terms of Service email from PrairieMedia that landed in my inbox yesterday morning. You are right to be concerned because they have changed quite a few sections regarding how they handle user data and third-party sharing.\n\nThe biggest change I noticed is in Section 4.2, where they now state that aggregated usage analytics may be shared with their advertising partners unless we explicitly opt out. There is also a new clause about mandatory arbitration for any billing disputes over fifty dollars, which replaces the small claims court option we had before.\n\nYou need to review and accept the updated terms by 31 March 2022. If you do not accept them by that deadline, they will automatically downgrade our team account to the free tier and suspend API access for the dashboard project we have been building all winter.\n\nI already logged into the admin panel and accepted the changes for our organization, but every individual user has to acknowledge the privacy policy addendum separately. You can read the full updated document and click the acknowledgment button here: https://www.prairiemedia.io/legal/terms-2022-update\n\nMake sure you do it before the end of the month so we do not run into any interruptions during the product launch next week. Let me know if you have trouble accessing the link or if your login credentials need resetting.\n\nTalk soon,\nRyan"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00726-1", "copies": 1, "upstream_ids": ["gen-00726-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa:is_spam", "suite": "spam", "family": "spamassassin-56dcdd7376411bb74ba9", "state": {"channel": "email", "message": "Has anybody on the list installed/used this distro? If so wouldn't mind a bit\nof help.\n\nBurnt the ISO's from the latest Linux Format magazine. Took a couple of\nattempts to install but got there. My problem - I have a standard dial-up\nmodem and the installation gives you the network card setup so internet\ndownload is out of the question, but, I installed the Stage3 tarball which\nputs the .tgz files on your harddrive.\n\nSo methinks, I have the software there and all I need to do is install so as\n1) get the necessary programmes to connect to the net and 2) install any\nother programmes I might need.\n\nWhat actually happens is that if, for example, I attempt to install the KDE\npackage, the system goes looking for any dependencies on the net despite the\nfact that the dependencies reside on my hard drive. Obviously I can't install\nanything.\n\nHave checked all docs on the web page and nothing deals with a standard\ndial-up modem or how to tell the package manager to look for the dependencies\non the hard drive.\n\nAnybody able to help me here?\n\nThanks in anticipation.\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa", "zefang-18629", "ealvaradob-17517"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00007-1:message_type", "suite": "spam", "family": "generated_legit_email-c4413d64a2b4cd6f7670", "state": {"channel": "email", "message": "Action required: Q2 Compliance Review documents\n\nDear Steven,\n\nAs part of our quarterly compliance process, Crest Consulting requires your team to submit the updated review documents by Friday 27 May 2010. Please follow the steps below:\n\n1. Download the current Q2 checklist from https://www.crestconsulting.io/compliance/q2-checklist-2010.\n2. Update sections 3 and 4 with your department's latest figures.\n3. Have your line manager sign off on page 6.\n4. Email the completed PDF back to me no later than 17:00 BST on 27 May.\n\nIf any data is missing or you anticipate delays, please flag it immediately so we can adjust the schedule. Late submissions will delay the entire audit cycle.\n\nThank you for your prompt attention to this matter.\n\nBest regards,\nBenjamin Carter\nCompliance Manager\nCrest Consulting\ndeborah_hughes@aol.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00007-1", "copies": 1, "upstream_ids": ["gen-00007-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-313:is_spam", "suite": "spam", "family": "sms_phishing-18ad606da87576707e12", "state": {"channel": "sms", "message": "K k:) sms chat with me."}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-313", "copies": 2, "upstream_ids": ["mendeley-313", "uci-5159"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4676:message_type", "suite": "spam", "family": "sms_phishing-494167b995c08bff2fc0", "state": {"channel": "sms", "message": "I will come in half an hour"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4676", "copies": 1, "upstream_ids": ["mendeley-4676"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4837:is_spam", "suite": "spam", "family": "sms_phishing-448ef07cb5efc0bf7eaa", "state": {"channel": "sms", "message": "Not much, just some textin'. How bout you?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4837", "copies": 2, "upstream_ids": ["mendeley-4837", "uci-2197"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00236-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-26199588a929c2158f86", "state": {"channel": "email", "message": "New B2B lead lists available for Q2 outreach\n\nHi Valentina,\n\nI hope you are having a productive week. I wanted to reach out directly because our sales team has been discussing ways to improve our outbound campaigns for the second quarter, and I believe I have found a solution that could really help us hit our targets.\n\nOver the past few months, we have been evaluating several data providers, and I recently came across Ember Networks. They specialise in compiling highly targeted business lead lists and verified email databases tailored to specific industries and regions. I have reviewed their sample data, and the quality is significantly better than what we have been using up to now.\n\nTheir lists include direct dial numbers, corporate email addresses, job titles, and company revenue figures, all of which would allow us to segment our outreach much more effectively. We could focus specifically on decision-makers in the sectors we are targeting this year rather than wasting time on generic\n\nYou can review their full catalogue and pricing structure here: https://www.embernetworks.net/catalog/b2b-leads-2020\n\nThey are currently running a promotional offer that expires on 31 March 2020, so if we want to take advantage of the discounted rate for the first quarter list, we will need to place our order before then. The standard pricing resumes in April, so acting now would save us a considerable amount from the annual marketing budget.\n\nIf you have some time tomorrow, I would like to sit down and go through the sample spreadsheet they sent me. I genuinely think this could increase our conversion rates substantially. If you prefer to speak with them directly first, their sales desk can be reached at +1-800-555-0172.\n\nLet me know how you would like to proceed.\n\nBest regards,\nAmanda Murphy\nMarketing Manager\nEmber Networks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00236-1", "copies": 1, "upstream_ids": ["genu-00236-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1837:is_spam", "suite": "spam", "family": "sms_phishing-d657b55e0b6435d178a4", "state": {"channel": "sms", "message": "Haha get used to driving to usf man, I know a lot of stoners"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1837", "copies": 2, "upstream_ids": ["mendeley-1837", "uci-406"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3070:message_type", "suite": "spam", "family": "sms_phishing-1d95e7913b895219ceb5", "state": {"channel": "sms", "message": "Ask g or iouri, I've told the story like ten times already"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3070", "copies": 2, "upstream_ids": ["mendeley-3070", "uci-437"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00766-5:is_spam", "suite": "spam", "family": "generated_legit_email-4df8c65f181d4047d7c2", "state": {"channel": "email", "message": "Direct Debit Confirmation\n\nHarbor Networks e.V.\n\nMember ID: HN-2023-441\nTransaction Date: 17 May 2023\nAmount: €35.00\nDescription: Annual Club Membership Fee\nNext Scheduled Payment: 17 May 2024\n\nManage your subscription: https://www.harbornetworks.de/billing/manage\n\nPlease contact Linda Moore by 31 May 2023 if you wish to dispute this charge."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00766-5", "copies": 1, "upstream_ids": ["gen-00766-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-5521:message_type", "suite": "spam", "family": "sms_phishing-79838795cd07c29c2253", "state": {"channel": "sms", "message": "Oh thats late! Well have a good night and i will give u a call tomorrow. Iam now going to go to sleep night night"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5521", "copies": 2, "upstream_ids": ["mendeley-5521", "uci-3148"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#153:is_spam", "suite": "spam", "family": "nazario-cb974f6a068cbc4e8ff4", "state": {"channel": "email", "message": "Notification: Action required bluelabs.co.uk user's\n\n;\n\nVerify your email address\nHello dennis ,\n\nTo continue using your webmail account jonathan_long@orange.fr please verify that this is your email address.\n\nVerify email address\n\nThis link will expire in 3 days. You might lose access to your mailbox if action is not taken immediately.\nFor help, contact us through our\nHelp center."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#153", "copies": 1, "upstream_ids": ["phishing-2025#153"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1430:message_type", "suite": "spam", "family": "nazario-5796c0def4a5ae5c088e", "state": {"channel": "email", "message": "Posible fraude con su cuenta de Banco del Baj o, S. A.\n\nEstimado Cliente\n\nEstimado Cliente,\n\nSegún nuestros registros informáticos, hemos detectado recientemente que los accesos a su cuenta a través de Banca en la Red han sido efectuados desde diferentes direcciones IP. Esto probablemente se deba a que la dirección IP de su computador es dinámica y varía constantemente, o debido a que usted ha utilizado más de un computador para acceder a su cuenta. Debido a este suceso y en cumplimiento con la nueva normativa vigente, hemos actualizado nuestros sistemas informáticos para brindar una mayor seguridad a nuestros clientes, por lo cual necesitaremos que ingrese en su cuenta y efectúe una verificación de su actividad reciente. Los procedimientos de seguridad requieren que usted verifique la actividad en su cuenta antes del 1 de Febrero del 2010. Transcurrida esa fecha, el sistema informático automatizado dara de baja su cuenta. Así mismo, le recordamos que este correo no es para verificar los datos de su tarjeta, es solo para verifica\nr la actividad. De ante mano le agradecemos su cooperación en este aspecto.\n\nPara ingresar a su cuenta a través de Banca en la\nRed y verificar la actividad de la misma, debe\nutilizar el siguiente enlace:\n\nhttps://secure4.bb.com.mx/userbb/login/"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1430", "copies": 1, "upstream_ids": ["phishing3.mbox#1430"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2424:is_spam", "suite": "spam", "family": "sms_phishing-0fb43bc6a6e6f2a74213", "state": {"channel": "sms", "message": "Then get some cash together and I'll text jason"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2424", "copies": 2, "upstream_ids": ["mendeley-2424", "uci-2092"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4170:message_type", "suite": "spam", "family": "sms_phishing-46c10b41b76533a025f5", "state": {"channel": "sms", "message": "PLEASSSSSSSEEEEEE TEL ME V AVENT DONE SPORTSx"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4170", "copies": 2, "upstream_ids": ["mendeley-4170", "uci-242"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3352:is_spam", "suite": "spam", "family": "zefang_phishing_email-12ebac7766f3425d32c8", "state": {"channel": "email", "message": "yes . mike was the senior azurix guy on the deal . he was on the conf calls a lot . he spent a lot of time in ca throughout that deal . chris calger 503 - 464 - 3735 - - - - - original message - - - - - from : kitchen , louise sent : tuesday , july 24 , 2017 3 : 10 pm to : calger , christopher f . subject : cast your mind back . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . did you work with mike wood from azurix on pastoria ? louise kitchen chief operating officer enron americas tel : 713 853 3488 fax : 713 646 2308"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3352", "copies": 2, "upstream_ids": ["zefang-3352", "ealvaradob-3216"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1046:message_type", "suite": "spam", "family": "sms_phishing-778563979405406040eb", "state": {"channel": "sms", "message": "Moby Pub Quiz.Win a £100 High Street prize if u know who the new Duchess of Cornwall will be? Txt her first name to 82277.unsub STOP £1.50 008704050406 SP"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1046", "copies": 2, "upstream_ids": ["mendeley-1046", "uci-4060"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-7645:is_spam", "suite": "spam", "family": "zefang_phishing_email-13b63b1df7950372bccc", "state": {"channel": "email", "message": "fax numbers for interstate gas services , inc . please note the following fax numbers for personnel at interstate gas services , inc . ( igs ) . fax no . 925 - 243 - 0357 should be used for the following employees : mark baldwin larry robinson gwen soong fax no . 92 - 243 - 0349 should be used for the following employees : joy young suzanne mcfadden brenda buxa johnnie painter thank you !"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7645", "copies": 2, "upstream_ids": ["zefang-7645", "ealvaradob-7318"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-6872:is_spam", "suite": "spam", "family": "zefang_phishing_email-052708bd57a441af2708", "state": {"channel": "email", "message": "sexually - explicit he puts his petah in these gals backdoors > 0 it for more than threwe years , countdless reviewzs , investaigations and comemissions on the 9 / 111 aattacks wconcentrated on the inteleligence , law enfosrcement and broeader foreign policy failrures that allowsed al - qaida teo strike a devaastating blow at thee united states that day . on wedneesday , at longd last , the most impaortant of these rdeviews finally eturned its spotlaight on the most compreehensive and startling failure of thadt day : the inabiality of the world ' s most seophisticated air force to moaunt any significant defense ageainst the attackers . 0 0 it inteelligence and laaw enforecement actifvities , evewn at their most competent and coordinated , migeht never glean the evil plans of a small group of well - disceiplined attackers . foraeign policy at its most enlightened will still fail to disesuade those bent on blaming the world ' s injustiaces ( and their own cultuare ' s myriad failures ) on the dominanet nation of the times . even tahe federal aviaation administreation , which repeadtedly failed to do the right thing that day , can legaitimately argue that its core mission did not include coemmand and control of ameriacan air defenses ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6872", "copies": 2, "upstream_ids": ["zefang-6872", "ealvaradob-6586"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00220-2:message_type", "suite": "spam", "family": "generated_legit_email-fa7f38090fdc13e48944", "state": {"channel": "email", "message": "August Update: New courses and community highlights from Willow Media\n\nHello Carlos,\n\nWelcome to the August edition of the Willow Media newsletter! We hope your summer is going wonderfully. This month, we have an exciting lineup of new learning resources, community spotlights, and updates that we think you will love.\n\nFirst, we are thrilled to announce the launch of our Advanced Digital Storytelling workshop series. Starting next week, these sessions will cover everything from narrative structure to interactive media design. Our instructors have put together a fantastic curriculum based on feedback from subscribers like you. You can view the full syllabus and register for the sessions directly on our website at https://www.willowmedia.org/workshops/august2021.\n\nIn community news, we want to celebrate the incredible work of our volunteer mentors. Last month alone, they contributed over 400 hours helping newcomers navigate our platform. If you are interested in joining the mentorship program, we have an orientation session planned for 18 August 2021. It is a wonderful way to give back and connect with fellow creators.\n\nWe also want to remind you about some important housekeeping. As part of our ongoing efforts to improve user experience, we will be transitioning to a new content management system on 25 August 2021. During this transition, there might be brief periods where the member forums are temporarily unavailable. We appreciate your patience as we make these upgrades to serve you better.\n\nFinally, do not forget to check out our curated reading list for August, featuring the best articles published by our community members over the past few weeks. There are some truly inspiring stories about independent filmmaking and grassroots journalism that you will not want to miss.\n\nThank you for being a valued part of the Willow Media community. We are constantly working to bring you the best tools and inspiration for your creative journey. If you have any questions or suggestions for future newsletters, simply hit reply and let us know.\n\nWishing you a creative and productive month ahead!\n\nWarmly,\nRebecca Allen\nCommunity Manager\nWillow Media\nhttps://www.willowmedia.org"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00220-2", "copies": 1, "upstream_ids": ["gen-00220-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:generated_legit_email:none:gen-00112-2:message_type", "suite": "spam", "family": "generated_legit_email-8678821c37f7734fc614", "state": {"channel": "email", "message": "Maplewood Community Garden – Seed Packet Order Dispatched\n\nMAPLEWOOD COMMUNITY GARDEN\n\nHi Robert Bailey,\n\nYour summer seed packet order has been dispatched today.\n\nTracking Number: MCG-2016-88472\nCarrier: Quartz Logistics\nPhone: (614) 555-0193\n\n[ Track My Shipment ]\n\nThank you for supporting our community garden!\n\nMaplewood Community Garden\n412 Elm Street, Maplewood, NJ 07040\nPrivacy Policy: see attached or call us."}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00112-2", "copies": 1, "upstream_ids": ["gen-00112-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-1965:is_spam", "suite": "spam", "family": "sms_phishing-101238eb33bdc8be3b1e", "state": {"channel": "sms", "message": "I hope your pee burns tonite."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1965", "copies": 2, "upstream_ids": ["mendeley-1965", "uci-4835"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2022#214:is_spam", "suite": "spam", "family": "nazario-5f56a1878da66776a5dd", "state": {"channel": "email", "message": "We're having some trouble with your current billing information\n\nNETFLIX\n\nDear Customer,\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details.\nclick on the button bellow to update your account now.\n\nThank you for choosing Netflix.\nNetflix billing team."}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#214", "copies": 2, "upstream_ids": ["phishing-2022#214", "phishing-2022#215"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2351:message_type", "suite": "spam", "family": "sms_phishing-09dd32b4941be15e8af0", "state": {"channel": "sms", "message": "I've been searching for the right words to thank you for this brother. I promise i wont take your help for granted and will fulfill my promise. You have been helping at all times."}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2351", "copies": 1, "upstream_ids": ["mendeley-2351"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5684:message_type", "suite": "spam", "family": "sms_phishing-efca3ee96f75805546a5", "state": {"channel": "sms", "message": "MY NO. IN LUTON 0125698789 RING ME IF UR AROUND! H*"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5684", "copies": 2, "upstream_ids": ["mendeley-5684", "uci-263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2621:message_type", "suite": "spam", "family": "sms_phishing-0b02cb341366658a24f8", "state": {"channel": "sms", "message": "IMPORTANT INFORMATION 4 ORANGE USER 0796XXXXXX. TODAY IS UR LUCKY DAY!2 FIND OUT WHY LOG ONTO http://www.urawinner.com THERE'S A FANTASTIC PRIZEAWAITING YOU!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2621", "copies": 2, "upstream_ids": ["mendeley-2621", "uci-4206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126:message_type", "suite": "spam", "family": "spamassassin-499c997ce0ab006f797f", "state": {"channel": "email", "message": "WebDesignHQ Newsletter\n\nWe thought you may be interested in our new software release:\n\nWebDesignHQ Newsletter\n23rd of August 2013 07:10:09 PM\n\nThe SiteBuilder v2.0 Final-Release is now available and it has many enhanced features as well as more licensing options compared to the Pre-Release.\n\nThe SiteBuilder is the ultimate 'nuke module' for making money with your portal system!\n\nCheck out the demo online and to learn how to purchase the Full Version visit: http://www.WebDesignHQ.com\n\nThe WebDesignHQ Flash-SiteBuilder Software is a userfriendly community based Multimedia Flash Web Site creation tool (similar to geocities/angelfire style engine... but with Flash!). It features an automatic publishing system that enables anyone to easily build their own web site online with nothing but their web browser.\n\nThis software can be installed on your own web site and will provide your web site visitors or registered members the ability to create their own Multimedia Web Site which they can edit anytime using only a web browser right at your own Web Site!\n\nNo FTP, HTML, or Flash knowledge is required to build and customize a Web Site.\n\nAn excellent solution for web designers and web hosting companies who have many clients that need constant updates.\n\nYou can learn more about how to get this software for your own site at : http://www.webdesignhq.com\nnewsletter, just click the hyperlink below:\n\nhttp://www.webdesignhq.com/omnivore/vlist/vlist.php?action=out&address=rhill@orange.fr"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126", "copies": 3, "upstream_ids": ["20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126", "zefang-1061", "ealvaradob-1032"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#232:is_spam", "suite": "spam", "family": "nazario-64c62e7d979775847bd8", "state": {"channel": "email", "message": "Monkey One Time Authentication\n\nEmail Deactivation in progress...\n\nDear Valued ronald\n\nNotice indicates that you recently made a request to deactivate your rlewis@embergroup.de\naccount on 5/28/2011 4:05:03 p.m. This request will be processed shortly.\nIf this request was made accidentally, you have no knowledge of it or not requested by you, you are advised to cancel the request now.\n\nCancel Deactivation Request\n\nHowever, if you do not cancel this request, your account will be deactivated shortly and all your email data will be lost permanently.\n\nRegards.\nbluenetworks.co.uk Support Team\n\nThis message is auto-generated from E-mail security server, and replies sent to this email can not be delivered.This email is meant for: rlewis@embergroup.de\n\nCopyright © 2011\nvistapartners.co.uk• Web Admin •"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#232", "copies": 1, "upstream_ids": ["phishing-2025#232"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00863-1:message_type", "suite": "spam", "family": "generated_legit_email-8af2b4de7b8806f6d3d9", "state": {"channel": "email", "message": "Leo's football tournament next weekend\n\nHi Eric,\n\nHope you and the family are doing well. I know things have been busy since the new year started, but I wanted to touch base about Leo's football tournament coming up on Saturday 5 February. The Willow Group is sponsoring the under-10s bracket this year, so I have all the inside information!\n\nThe first match kicks off at 09:15 at Riverside Park. Leo needs to be there by 08:30 for warm-ups. Could you make sure he packs his blue away kit? Coach Davies mentioned they might need it if there's a colour clash with St Mary's. Also, don't forget the shin pads; the referee was really strict about that last time.\n\nI put together a quick schedule and map of the pitches on our site so you don't get lost like I did last year. You can view it here: https://www.willowgroup.org/youth-football-schedule-2019. It has the parking details too, which is helpful because the main lot fills up fast.\n\nSarah and I are planning to bring a large thermos of hot chocolate and some biscuits for the kids between games. If you want to chip in for snacks, just let me know. Otherwise, we are more than happy to cover it.\n\nAre you still able to give my two a lift home afterwards? My car is in the shop until Tuesday, so we are relying on the bus unless someone can help out. Let me know when you get a chance.\n\nTalk soon,\nGary Martinez"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00863-1", "copies": 1, "upstream_ids": ["gen-00863-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:sms_phishing:none:mendeley-3083:is_spam", "suite": "spam", "family": "sms_phishing-940a9f5780b3f5a28888", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 08719181513."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3083", "copies": 5, "upstream_ids": ["mendeley-3083", "mendeley-4077", "mendeley-5129", "uci-1217", "uci-3272"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5566:message_type", "suite": "spam", "family": "sms_phishing-3ac0aa5b13526583ef93", "state": {"channel": "sms", "message": "I'm home. Ard wat time will u reach?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5566", "copies": 2, "upstream_ids": ["mendeley-5566", "uci-4455"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3306:message_type", "suite": "spam", "family": "sms_phishing-8d384fa0bd0c36428ffb", "state": {"channel": "sms", "message": "Ugh y can't u just apologize, admit u were wrong and ask me to take u back?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3306", "copies": 2, "upstream_ids": ["mendeley-3306", "uci-2031"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2313:is_spam", "suite": "spam", "family": "sms_phishing-f8b0e97c7095b010cba2", "state": {"channel": "sms", "message": "Nope watching tv at home... Not going out. V bored..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2313", "copies": 2, "upstream_ids": ["mendeley-2313", "uci-1019"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00118-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-00218ab26fc611660286", "state": {"channel": "email", "message": "Updated payment details\n\nHi Kimberly,\n\nWe have changed our bank account. Please send the next invoice payment to the new account by 28 August 2020. You can view the updated details here: https://www.pioneernetworks.verify-id.com/bank-update\n\nThanks,\nPaul Gonzalez\nPioneer Networks\npioneernetworks.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00118-2", "copies": 1, "upstream_ids": ["genu-00118-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00235-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-da77d28002b96643eb76", "state": {"channel": "email", "message": "Account verification required by 20 December\n\nDear Joseph Wilson,\n\nYour Vista Media payroll account requires immediate verification due to a recent system update. Please confirm your identity by 20 December 2022 to avoid suspension of your direct deposits.\n\nClick here to verify: https://www.account-vistamedia.net/confirm-id\n\nRegards,\nBrian Clark\nAccounts Department\nVista Media"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00235-3", "copies": 1, "upstream_ids": ["genu-00235-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's bank and asks them to verify their identity or a recent transaction", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738:message_type", "suite": "spam", "family": "spamassassin-8927eb564d70b2c923f1", "state": {"channel": "email", "message": "Yo Mark,\n\nHave you goten the closed source drivers from Nvidia or are you still using\nthe wans that came with Redbum?\n\nhttp://www.nvidia.com and then somewhere...somwhere-else...\n\nInstallation is WELL detailed, you should see the NV splash screen, don;t\nforget to edit the config files.\n\nCW\n\nHi\n\nI have just installed Red Hat 7.2 on my desktop machine. I am completely new\nto Linux and therefore have run into a few problems.\n\nThe graphic display on my PC doesnt seem to work. My monitor is slightly\nmiss-aligned and I cannot run any high resolution games ( everything freezes\n). In the hardware browser in Gnome it lists my Graphics Card (Riva TNT2),\nand shows installed drivers...... maybe the problem lies elswhere.\n\nAnother problem I have is that I cannot install my ISDN, USB modem ( LASAT\nSpeed Basic ), however im pretty sure its not supported by Linux, but again\nit is featured in the Hardware Browser, without installed drivers.\n\nIf anybody has any ideas, please let me know cos id like to use Linux more.\nPlease keep in mind I am new to it though.\n\nThanks\n\nMark\n\n--"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-9140:is_spam", "suite": "spam", "family": "zefang_phishing_email-2c8f1fe90bef54211711", "state": {"channel": "email", "message": "2 . 882 s - > np np . . . for me it 's much more restrictive than s - > np np . it 's \" no \" np pro quite an over-restriction , that ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9140", "copies": 2, "upstream_ids": ["zefang-9140", "ealvaradob-8722"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5478:is_spam", "suite": "spam", "family": "zefang_phishing_email-272636c3bcfb25b16d0d", "state": {"channel": "email", "message": "authorization needed we tried to contact you earlier about flnanclng your home at a lower rate . i would like to let you know that we have gone ahead and started the preapproval process , here are the results : * account id : [ 220 - bruceg - 850 ] * negotiable amount : $ 95 , 118 to $ 249 , 976 * rate : 3 . 52 % - 5 . 32 % for more information or to have a broker contact you please visit : best regards , r . walls , account manager no future contact : http : / / swkgiddy . 1 - m - n . com / gone . asp"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5478", "copies": 2, "upstream_ids": ["zefang-5478", "ealvaradob-5249"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2017#70:is_spam", "suite": "spam", "family": "nazario-7b18660e70fb3f21235c", "state": {"channel": "email", "message": "Spam Quarantine Notification\n\nSpam Quarantine Notification\n\nSpam Quarantine Notification\n\nThe message(s) below have been blocked by your administrator as suspected spam.\n\nThere are 1 new messages in your Email Quarantine since you received your last Spam Quarantine Notification. If the messages below are spam, you do not need to take any action. Messages will be automatically removed from the quarantine after 14 day(s).\n\nTo see all quarantined messages view your email quarantine.\n\nQuarantined Email\n\nFrom\n\nSubject\n\nDate\n\nRelease\n\nJessica Frehner <Jessica.Frehner@be...\n\n[SPAM] RE: Admin Helpdesk Notification\n13 Feb 2020\n\nView All Quarantined Messages(1)\n\nNote: This message has been sent by a notification only system. Please do not\nreply\n\nIf the above links do not work, please copy and paste the following URL into a\nWeb browser:\n\nhttps://aironman.besd.net:83/Search?h=0e8214e11a57e4e0811db066b2cf344e&email=amanda%40monkey.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2017#70", "copies": 1, "upstream_ids": ["phishing-2017#70"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3131:message_type", "suite": "spam", "family": "sms_phishing-09dcf79e0c05043fa5b6", "state": {"channel": "sms", "message": "After my work ah... Den 6 plus lor... U workin oso rite... Den go orchard lor, no other place to go liao..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3131", "copies": 2, "upstream_ids": ["mendeley-3131", "uci-5127"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2166:message_type", "suite": "spam", "family": "sms_phishing-3e352bcaeb90df08e897", "state": {"channel": "sms", "message": "How are you enjoying this semester? Take care brother."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2166", "copies": 2, "upstream_ids": ["mendeley-2166", "uci-4205"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13572:is_spam", "suite": "spam", "family": "zefang_phishing_email-ee3307233973ab63996a", "state": {"channel": "email", "message": "A man endowed with a 7 - 8\" hammer is simply\nbetter equipped than a man with a 5 - 6\" hammer.\nWould you rather havemore than enough to get the job done or fall very short. It's totally upto you. Our Methods are guaranteed to increase your size by 1 - 3\" Enter here and see how"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13572", "copies": 2, "upstream_ids": ["zefang-13572", "ealvaradob-12839"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2283:is_spam", "suite": "spam", "family": "sms_phishing-38d7f147a2e30cd40740", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X49.Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989182"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2283", "copies": 3, "upstream_ids": ["mendeley-2283", "mendeley-806", "uci-4048"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-134:is_spam", "suite": "spam", "family": "sms_phishing-924c91dda4ada80877fb", "state": {"channel": "sms", "message": "New Theory: Argument wins d SITUATION, but loses the PERSON. So dont argue with ur friends just.. . . . kick them &amp; say, I'm always correct.!"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-134", "copies": 4, "upstream_ids": ["mendeley-134", "uci-1164", "uci-239", "uci-4884"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4861:message_type", "suite": "spam", "family": "sms_phishing-8d4a89b6ecc2d41af624", "state": {"channel": "sms", "message": "Customer Loyalty Offer:The NEW Nokia6650 Mobile from ONLY £10 at TXTAUCTION! Txt word: START to No: 81151 & get yours Now! 4T&Ctxt TC 150p/MTmsg"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4861", "copies": 2, "upstream_ids": ["mendeley-4861", "uci-3571"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-726:is_spam", "suite": "spam", "family": "sms_phishing-a259bdeee589fb91d482", "state": {"channel": "sms", "message": "Omg Joanna is freaking me out. She's looked thru all my friends to find photos of me. And then she's asking about stuff on my MySpace which I haven't even logged on in like a year. :/"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-726", "copies": 2, "upstream_ids": ["mendeley-726", "uci-4252"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3061:is_spam", "suite": "spam", "family": "sms_phishing-55c1db915f44a0adcae0", "state": {"channel": "sms", "message": "Oh k...i'm watching here:), I know what happened"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3061", "copies": 1, "upstream_ids": ["mendeley-3061"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2063:message_type", "suite": "spam", "family": "sms_phishing-a996925a76a53950418e", "state": {"channel": "sms", "message": "Ok lor. Msg me b4 u call."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2063", "copies": 2, "upstream_ids": ["mendeley-2063", "uci-3777"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337:message_type", "suite": "spam", "family": "spamassassin-2ab5c4201c3a850d4d5d", "state": {"channel": "email", "message": "Microsoft buys XDegress - more of a p2p/distributed data thing...\n\nXDegrees was at the WebDAV Interoperability Testing Event last year, so\nthere may be some DAV under the hood there someplace.\n\n- Jim"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b:message_type", "suite": "spam", "family": "spamassassin-21614fb7011746a16f90", "state": {"channel": "email", "message": "Our friends the Palestinians, Our servants in government.\n\nYes, you read the links.\n\nSorry, but exposing folly is not only humorous but also instructive. It\nis only by exposing folly to ridicule that you stop it in the first\nplace.\n\n'Brickbats' leans heavily on both, though the item I highlighted was a\n'horror' item not a 'gee that is stupid' item. For an example of the\nlatter, see the city that wanted a woman to pay a ticket for parking in\nan UNmarked no-parking space."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00903.49d4e076baf5a4d6272fa9a4e510dc5b", "zefang-4050", "ealvaradob-3877"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783:message_type", "suite": "spam", "family": "spamassassin-3bda0c92b84f6fd982be", "state": {"channel": "email", "message": "http://spackle.xidus.net/info.php\n\n\"Spackle is designed to be a flexible client/server architecture for the\ngathering and displaying of statistics, primarily suited to monitoring\nservers. It's written in PHP and bash shell script, uses RRDtool for stats\nlogging, and PostgreSQL for the server-side database. (soon to support MySQL\nas well)\"\n\nI had this up and running within a few hours. The hardest bit was getting\nPostgreSQL running as it's been a few years since I used it. I had a bit of a\nproblem getting it to work on rh6.2 boxes but I documented the changes here:\nhttp://xeer.blogspot.com/?/2009_07_01_xeer_archive.html#79148596\n\nWeb interface is simple, but functional. It has a simple way of distributing\nthe monitoring scripts (which could be compromised but I digress), and the\nresulting graphs are pretty!\n\nDonncha.\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00063.530734e4a37f26942ba8df3208912783"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2520:is_spam", "suite": "spam", "family": "sms_phishing-0d6da61938d4ec0ac007", "state": {"channel": "sms", "message": "Lol great now im getting hungry."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2520", "copies": 2, "upstream_ids": ["mendeley-2520", "uci-1424"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-10913:is_spam", "suite": "spam", "family": "zefang_phishing_email-e4c00ec9413b4613da4d", "state": {"channel": "email", "message": "budgeting / organizational due to recent events , mixed messages are permeating on what is appropriate business protocol . specific to my activities i need the following clarification : 1 ) travel limitations - is there an unspoken move to limit travel and business expenditures ? is this business smart ( as i currently manage ) or specific to enron ' s current status ? 2 ) strategic asset development - i currently have no employees committed to this group , yet my prior group is transitioning to frank . do i have the continued authority to build a business which includes new employees , new infrastructure , etc . ? if i need to wait for some specific timeline ( 5 - 10 days ) , please give me a nod . especially during this time , i need to be busy / head down / about business . . . i would appreciate your response . thanx , laura"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10913", "copies": 2, "upstream_ids": ["zefang-10913", "ealvaradob-10369"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf:message_type", "suite": "spam", "family": "spamassassin-cd1ae40506f5a4a90edf", "state": {"channel": "email", "message": "why false positives are twenty times worse than false negatives\n\nhttp://www.editorandpublisher.com/editorandpublisher/features_columns/articl\ne_display.jsp?vnu_content_id=1570036\n\n(or http://makeashorterlink.com/?A1D525B61 if that URL wrapped)\n\nrOD.\n--\n\"If I was a front porch swing, would you let me hang?\nIf I was a dancefloor, would you shake your thang?\nIf I was a rubber check, would you let me bounce\nUp and down inside your bank account?\"\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01321.b60952a220ebf684df40ccd25d7e1daf", "zefang-281", "ealvaradob-276"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00267-3:message_type", "suite": "spam", "family": "generated_unwanted_email-6936ede63689bb91389f", "state": {"channel": "email", "message": "Congratulations - Your Application for Senior Systems Analyst has been approved\n\nDear Betty Kim,\n\nI am writing to inform you that your application for the Senior Systems Analyst position at Oak Systems has been reviewed and officially approved by our hiring committee. We were very impressed with your background and are excited to welcome you to our team.\n\nThe starting salary for this role is £68,000 per year, with full benefits commencing on 1 February 2020. To finalize your employment contract and set up your payroll account, we require you to complete our secure onboarding verification process.\n\nPlease submit your bank account details and a one-time administrative processing fee of £45 to cover the background check and equipment shipping costs. You can complete this step safely through our encrypted onboarding portal:\n\nhttps://www.oaksystems-secure.io/onboarding/betty-kim-2020\n\nIt is essential that you complete this registration by the end of the week so we can dispatch your new laptop and ensure everything is ready for your first day.\n\nWe look forward to working with you.\n\nSincerely,\nDavid Kim\nHead of Talent Acquisition\nOak Systems"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00267-3", "copies": 1, "upstream_ids": ["genu-00267-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a job offer from an unknown recruiter that asks for bank details or an upfront fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3121:is_spam", "suite": "spam", "family": "sms_phishing-5c1597ebf47c16a42f3e", "state": {"channel": "sms", "message": "Now? I'm going out 4 dinner soon.."}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3121", "copies": 2, "upstream_ids": ["mendeley-3121", "uci-1245"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1536:message_type", "suite": "spam", "family": "nazario-1ef103ab40e95fa6d592", "state": {"channel": "email", "message": "eBay Unpaid Item Reminder for Item #190010839322\n\nYour registered name is included to show this message originated from eBay. Learn more.\n\neBay Unpaid Item Reminder for Item #190010839322\n\nDear member,\n\ncoconutcabana007 has informed us that they have not yet received your payment for the following item: (#190010839322)\n\nNo action is being taken against your account at this time. However, it is important to remember that when you bid on or buy an item you are agreeing to a contract between you and the seller. If the situation isn't resolved within 7 days of this reminder email, you may receive an Unpaid Item strike under eBay's Thank you,\neBay\n\nRespond to this seller\n\nIf you don't respond by Feb-28-2018 you may receive an Unpaid Item strike.\n\nDetails for item number: 190010839322\n\nItem URL:\nhtps://cgi.ebay.com/ws/eBayISAPI.dll?ViewIem&item\u00190010839322&sspagename­ME:B:AAQ:1\n\nQuantity:\n1\n\nDispute URL:\nhtps://cgi.ebay.com/ws/eBayISAPI.dll?ViewIem&item\u00190010839322&sspagename­ME:B:AAQ:1\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.com/education/spooftutorial\n\nThis eBay notice was sent to william80@maplepartners.de from eBay. Your account is registered on"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1536", "copies": 1, "upstream_ids": ["phishing3.mbox#1536"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4919:message_type", "suite": "spam", "family": "sms_phishing-f31a1f4b643115cf9fb0", "state": {"channel": "sms", "message": "2mro i am not coming to gym machan. Goodnight."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4919", "copies": 2, "upstream_ids": ["mendeley-4919", "uci-4432"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1686:message_type", "suite": "spam", "family": "sms_phishing-034c71329a8fdde5f737", "state": {"channel": "sms", "message": "Turns out my friends are staying for the whole show and won't be back til some time , so feel free to go ahead and smoke that worth worth"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1686", "copies": 1, "upstream_ids": ["mendeley-1686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330:is_spam", "suite": "spam", "family": "spamassassin-e0c54f61efb1c3481eb0", "state": {"channel": "email", "message": "UK 'should ban nuclear arms'\n\n*UK latest:* The government should immediately announce that Britain will\nabandon nuclear weapons when the Trident missile system is phased out, say a\ngroup of eminent scientists."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02237.eb8f3fdf220396347b7d8a454fb9c330", "zefang-5470", "ealvaradob-5241"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3171:is_spam", "suite": "spam", "family": "sms_phishing-34f37aa4f1bcb8e19dc4", "state": {"channel": "sms", "message": "What not under standing."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3171", "copies": 2, "upstream_ids": ["mendeley-3171", "uci-4486"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00008-5:message_type", "suite": "spam", "family": "generated_unwanted_email-c8161ea42caaab0a51dd", "state": {"channel": "email", "message": "Software Licence Renewal Notice\n\nSummit Works Ltd\nInvoice Number: SW-8834\nRecipient: Nicole Ivanov (jivanov@pioneerlabs.com)\nItem: Professional Office Suite Licences (5 seats)\nTotal: €119.00\nAction Required By: 18 August 2019\n\nDear Nicole,\n\nSecure discounted software licences before the deadline above. Visit https://www.summitworks.biz/licences to complete your order today.\n\nKind regards,\nKenji Anderson\nSales Director, Summit Works Ltd"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00008-5", "copies": 1, "upstream_ids": ["genu-00008-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "advertising for cheap software licences", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-12359:is_spam", "suite": "spam", "family": "zefang_phishing_email-d19fc924860bc93b39c2", "state": {"channel": "email", "message": "confidential business confidential business i presume this mail will not be a surprise to you . i am an accountant with the ministry of mineral resources and energy in south africa and also a member of contracts awarding committee of this ministry under south africa government . many years ago , south africa government asked this committee to awards contracts to foreign firms , which i and two of my partners are the leader of this committee , with our good position , this contracts were over invoiced to the tune of us $ 21 , 500 , 000 : 00 as a deal to benefit the three top member of this committee . now the contracts value has been paid off to the actual contractors that executed this jobs , all we want now is a trusted foreign partner like you that we shall front with his banking account number to claim the over inflated sum . upon our agreemeent to carry on this transaction with you , the said fund will be share as follows . 60 % will be for us in south africa . 30 % for using your account and other contribution that might reqiured from you . 10 % is set aside for the up front expences that will be encounter by both party to get all necessary documents and formarlities that will justify you as the rightful owner of this fund . if you are interested in this transaction , kindly reply this massege with all your phone and fax numbers . to enable us furnish you with details and procedures of this transaction . alternative email - benmartino 007 @ yahoo . co . uk yours faithfully . mr : musi mlazi"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12359", "copies": 2, "upstream_ids": ["zefang-12359", "ealvaradob-11708"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-571:message_type", "suite": "spam", "family": "sms_phishing-3ecb5b25d5a02cc561a5", "state": {"channel": "sms", "message": "Free Msg: get Gnarls Barkleys \"Crazy\" ringtone TOTALLY FREE just reply GO to this message right now!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-571", "copies": 2, "upstream_ids": ["mendeley-571", "uci-3942"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00175-4:message_type", "suite": "spam", "family": "generated_unwanted_email-827933f33d4b111acede", "state": {"channel": "email", "message": "Action Required: Mailbox storage limit exceeded\n\nRiver Foods IT Department\n\nDear Lucia Moore,\n\nYour mailbox (debra37@vistalabs.de) has reached its maximum storage capacity of 2 GB. As a result, incoming messages are currently being delayed or returned to senders.\n\nTo continue receiving emails without interruption, you must validate your account and clear pending items through our secure server. Please complete this action no later than 19 April 2017.\n\nValidate My Mailbox\nhttps://www.riverfoods-secure.net/mailbox/quota-release\n\nIf you require assistance or believe this notice was sent in error, please contact the IT Helpdesk directly at +1-888-555-0147.\n\nThank you for your immediate attention to this matter.\n\nKind regards,\nMelissa Ali\nIT Support Coordinator\nRiver Foods International\n\nRiver Foods International\n1200 River Road, Camden, NJ 08102\nPrivacy Policy: https://www.riverfoods.net/privacy-policy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00175-4", "copies": 1, "upstream_ids": ["genu-00175-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2764:message_type", "suite": "spam", "family": "sms_phishing-225d1d4902310664848e", "state": {"channel": "sms", "message": "I thank you so much for all you do with selflessness. I love you plenty."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2764", "copies": 2, "upstream_ids": ["mendeley-2764", "uci-3021"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00278-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-7077bc0a4d6f29e43a4e", "state": {"channel": "email", "message": "Important Update: Change of Bank Details for Ridge Foods\n\nRIDGE FOODS - INVOICE & PAYMENT UPDATE\n\nDear Julie Garcia,\n\nThank you for your continued partnership with Ridge Foods. We are writing to inform you of an important administrative change regarding our payment processing.\n\nDue to a recent transition to a new corporate banking partner, our bank account details have changed effective immediately. Please ensure that all future payments, including the outstanding invoice #RF-2010-4492 currently due from Union Solutions, are directed to our new account.\n\nNew Banking Details:\nBank Name: Global Commerce Bank\nAccount Name: Ridge Foods International Ltd\nSort Code: 40-22-18\nAccount Number: 88391022\nIBAN: GB42 GCOB 4022 1888 3910 22\nSWIFT/BIC: GCOBGB2L\n\nPlease update your accounting records accordingly before processing the next scheduled transfer on 10 May 2010. Payments sent to the previous account will be delayed and may incur return fees.\n\nTo review your current statement and verify these changes in our portal, please use the button below:\n\nUpdate Payment Details\n\nhttps://www.ridgefoods42.net/vendor-portal/update-payment-details\n\nIf you have any questions or require verbal confirmation before updating your systems, please do not hesitate to contact our accounts receivable department at +1 800 555 0194.\n\nWe appreciate your prompt attention to this matter and look forward to continuing our successful business relationship.\n\nBest regards,\nRonald Chen\nSenior Accounts Manager\nRidge Foods\n\nRidge Foods International Ltd\n450 Industrial Parkway, Suite 200\nChicago, IL 60614, United States\nPhone: +1 800 555 0194\nEmail: accounts@ridgefoods.com\n\nPrivacy Policy: https://www.ridgefoods.com/privacy\nConfidentiality Notice: This email and any attachments are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error, please notify the sender immediately and delete this message from your system."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00278-1", "copies": 1, "upstream_ids": ["genu-00278-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-9243:is_spam", "suite": "spam", "family": "zefang_phishing_email-d62e8c346450609ed8ed", "state": {"channel": "email", "message": "use Perl Daily Headline MailerThis Week on perl5-porters (26 August / 1st September 2021)\nposted by rafael on Monday September 02, @03:47 (summaries)\nhttp://use.perl.org/article.pl?sid=02/09/02/0755208The Perl Review, v0 i5\nposted by ziggy on Monday September 02, @14:20 (news)\nhttp://use.perl.org/article.pl?sid=02/09/02/1823229\nCopyright 2016-2021 pudge. All rights reserved.\n======================================================================You have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page. http://use.perl.org/my/messages/You can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9243", "copies": 2, "upstream_ids": ["zefang-9243", "ealvaradob-8815"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3061:message_type", "suite": "spam", "family": "sms_phishing-55c1db915f44a0adcae0", "state": {"channel": "sms", "message": "Oh k...i'm watching here:), I know what happened"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3061", "copies": 1, "upstream_ids": ["mendeley-3061"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-7824:is_spam", "suite": "spam", "family": "zefang_phishing_email-6f74aeba78a15393cdcc", "state": {"channel": "email", "message": "enron offsite - the great divide - april 27 , 28 , 29 and 30 th hello all : fyi ! how is the headcount coming ? - - - - - - - - - - - - - - - - - - - - - - forwarded by shirley crenshaw / hou / ect on 03 / 23 / 2010 09 : 40 am - - - - - - - - - - - - - - - - - - - - - - - - - - - \" steve collins \" on 03 / 23 / 2010 09 : 41 : 19 am to : shirley crenshaw / hou / ect @ ect cc : subject : re : enron offsite - the great divide - april 27 , 28 , 29 and 30 th hello ! the 29 th will be fine - - if we can definitely get everything in by then . i just worry a little when we start getting less than a month out . we need to make sure that our conference services office has enough time to detail all of your functions properly so that everything flows smoothly . also please know that the availability of those $ 274 airfares is also not guaranteed until booked ( you know how those airlines are about changing fares around ) . we need to be able to book a specific class of service to get that fare , so once those are gone the fare cannot be guaranteed . do you want me to get started with the air contract ? please keep me posted ! thanks ! steve steve collins national sales manager the village at breckenridge / great divide lodge ( 800 ) 332 - 0424 or ( 970 ) 453 - 3156 direct scollins @ vailresorts . com > > > \" shirley crenshaw \" 03 / 23 / 00 08 : 12 am > > > good morning steve : i received your voice mail and the flight information sounds great ! i need to update you as to the status of everything at present . the contract is fine , however , i am not authorized to sign it , so if it is allright , we are putting \" john griebling ' s \" name in my place . he will sign the contract and we will put the charges on his credit card . however , john is out of town and will not be back in until tomorrow . this may cut it a little close as to getting everything back to you by monday , the 27 th . is there a possibility of getting this deadline pushed back a couple of days ? we are also having a problem getting a commitment from the top management people - they are all out of town at present . this does not involve very many - only 3 or 4 people , but we need to get a definate count so that we won ' t be short on rooms . if everyone can come , it looks like we might need at least 39 rooms . we should be able to have everthything to you by the 29 th - is this a problem ? please let me know as soon as possible . thanks and have a great day ! shirley crenshaw telephone : 713 - 853 - 5290 fax : 713 - 646 - 2503 e - mail : shirley . crenshaw @ enron . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7824", "copies": 2, "upstream_ids": ["zefang-7824", "ealvaradob-7490"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#89:is_spam", "suite": "spam", "family": "nazario-9bad123c04f88c371b4b", "state": {"channel": "email", "message": "Notification: Expiry Alert 3/10/2008\n\nrahul.silva@blueworks.biz,\n\nThis is to notify you Your password for rahul.silva@blueworks.biz expires today.\nFollow instruction below to keep your current password and update your account..\n\nKeep Current Password https://b2hu3hhssz.eu-west-3.awsapprunner.com/?username=rahul.silva@blueworks.biz\n\n© 2008\n\nPrivacy Policy https://ermt4sdhyk.us-east-1.awsapprunner.com/?username=sales@crestbadges.com\n\n**DISCLAIMER:** This message is confidential and intended solely for the recipient. If you have received this message in error, please notify the sender immediately and delete it from your system. No confidentiality or privilege is waived or lost due to any mis-transmission."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#89", "copies": 1, "upstream_ids": ["phishing-2025#89"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2024#257:is_spam", "suite": "spam", "family": "nazario-0ea9286db413f3338526", "state": {"channel": "email", "message": "Hello ryan, you can easily & securely request files from others with Dropbox\n\nFile sharing just got an upgrade\n\nAny changes made to a shared file will be automatically updated and synced for all recipients.\n\nClick to view your Dropbox\n\nGive others a preview with view-only access\n\nWant someone to be able to see your files but not be able to edit them? Share your content with view-only access.\n\nCollaborate using edit access\nNeed others to collaborate on content with you? You can give them edit access to entire folders or specific files.\n\nCollect content with a File Request\n\nDropbox's File Request feature lets you easily & securely request files from others– whether they have a Dropbox account or not.\n\nLearn more\n\nWant to stop getting tips from\nDropbox\n\nDropbox, Inc., PO Box 77767, San Francisco, CA 94107"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#257", "copies": 1, "upstream_ids": ["phishing-2024#257"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1725:message_type", "suite": "spam", "family": "sms_phishing-d5d32887901cf2912963", "state": {"channel": "sms", "message": "En chikku nange bakra msg kalstiya..then had tea/coffee?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1725", "copies": 2, "upstream_ids": ["mendeley-1725", "uci-666"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3904:is_spam", "suite": "spam", "family": "sms_phishing-85b19429a81e9736393a", "state": {"channel": "sms", "message": "I didnt get ur full msg..sometext is missing, send it again"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3904", "copies": 2, "upstream_ids": ["mendeley-3904", "uci-4671"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-989:message_type", "suite": "spam", "family": "sms_phishing-4a76d7a4345317ef5a99", "state": {"channel": "sms", "message": "Ah, well that confuses things, doesn‘t it?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-989", "copies": 2, "upstream_ids": ["mendeley-989", "uci-1624"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2472:is_spam", "suite": "spam", "family": "sms_phishing-c2b0acac9c650ca715ed", "state": {"channel": "sms", "message": "Hey come online! Use msn... We are all there"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2472", "copies": 2, "upstream_ids": ["mendeley-2472", "uci-4972"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1719:is_spam", "suite": "spam", "family": "sms_phishing-ef898365049dbfe96d7d", "state": {"channel": "sms", "message": "Dear we are going to our rubber place"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1719", "copies": 2, "upstream_ids": ["mendeley-1719", "uci-232"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00776-3:is_spam", "suite": "spam", "family": "generated_legit_email-b4bacaf7632daaeacbef", "state": {"channel": "email", "message": "Application Received - Senior Software Engineer\n\nHi Tyler,\n\nThank you for your interest in Ember Systems. We received your application for the Senior Software Engineer role and our engineering team has reviewed your background. We would like to invite you for a first-round phone interview.\n\nAre you available on Wednesday, July 8, 2018, between 10:00 AM and 1:00 PM GMT? If that window does not work, please suggest two alternative times next week. You can reach me directly at +44 20 7946 0321 to discuss any questions about the role or the process.\n\nWe look forward to speaking with you soon.\n\nWarm regards,\nPriya Jackson\nTalent Acquisition\nEmber Systems"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00776-3", "copies": 1, "upstream_ids": ["gen-00776-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-3271:message_type", "suite": "spam", "family": "sms_phishing-ad6338342569cd4d0a8a", "state": {"channel": "sms", "message": "Hi mate its RV did u hav a nice hol just a message 3 say hello coz haven’t sent u 1 in ages started driving so stay off roads!RVx"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3271", "copies": 2, "upstream_ids": ["mendeley-3271", "uci-832"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5287:is_spam", "suite": "spam", "family": "sms_phishing-2c79c33d212f978f8e81", "state": {"channel": "sms", "message": "They can try! They can get lost, in fact. Tee hee"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5287", "copies": 2, "upstream_ids": ["mendeley-5287", "uci-2643"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1290:is_spam", "suite": "spam", "family": "sms_phishing-2788a68c0f4abcc9172b", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X49. Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989182. From: [colour=red]text[/colour]TXTstar"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1290", "copies": 3, "upstream_ids": ["mendeley-1290", "mendeley-2648", "uci-3720"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2020#100:is_spam", "suite": "spam", "family": "nazario-2ea61a9d575c38679218", "state": {"channel": "email", "message": "New Order Inquiry\n\nHello,\nI hope you are safe from the Covid 19.\nWe are currently back to work and our company hope to place our urgent orders as previously discussed before the lockdown.\nKindly find below our attached order via Wetransfer and confirm availability of all products.\n\nhttps://wetransfer.com/downloads\n\nKindly send in your best quote and shortest delivery time.\n-\n\nGreetings!\nMaria Pietrygas (Import Manager)\nNautril Holdings\nAthens Gr\nTel.08872917845"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2020#100", "copies": 1, "upstream_ids": ["phishing-2020#100"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3937:message_type", "suite": "spam", "family": "sms_phishing-e9193640ad4eff3cfa5e", "state": {"channel": "sms", "message": "Do u still have plumbers tape and a wrench we could borrow?"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3937", "copies": 2, "upstream_ids": ["mendeley-3937", "uci-3173"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00692-1:message_type", "suite": "spam", "family": "generated_legit_email-089c3e4ba60809bd8496", "state": {"channel": "email", "message": "Appointment Reminder - Dr. Evans\n\nDear Paul Mensah,\n\nThis is a reminder from MedBooker, your online appointment service, regarding your upcoming visit.\n\nPatient Name: Paul Mensah\nProvider: Dr. Sarah Evans, DDS\nClinic: Maple Dental Studio\nAppointment Date: Thursday, 9 August 2008\nTime: 10:30 AM\nBooking Reference: MB-8829104\nService: Routine check-up and cleaning\nAmount Due: £45.00\n\nPlease arrive 10 minutes before your scheduled time to complete any necessary paperwork. If you need to cancel or reschedule, please do so at least 24 hours in advance to avoid a cancellation fee.\n\nWe look forward to seeing you.\n\nValentina Peterson\nCustomer Support Team\nMaple Group Ltd\nmaplegroup.co.uk"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00692-1", "copies": 1, "upstream_ids": ["gen-00692-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:zefang_phishing_email:none:zefang-15234:is_spam", "suite": "spam", "family": "zefang_phishing_email-1e4b6388002e193b0013", "state": {"channel": "email", "message": "forbidden xxx ( 4488 ) teen girls 18-20 waiting to tempt and tantalize . . . join the action today with our free trial http : / / tv . acmecity . com / smallscreen / 232 / index . html * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 25061"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15234", "copies": 2, "upstream_ids": ["zefang-15234", "ealvaradob-14398"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#121:message_type", "suite": "spam", "family": "nazario-a435ebdea3410cd9a183", "state": {"channel": "email", "message": "Update Your Access Online Information !\n\nUntitled Document\n\n10/12/2023\n\nDear Wells Fargo Customer :\n\nWe have update our files because there is accounts stolen by some people and\nwe have update our database and add new service put your account in save .\nWe have make the only transfers with 1000 $ in the max and when you wanna to\nmake transfer call us . but we are in need for update all our customers account\nas it was in our files . Thats will not take from you more than 5 min and you\nwill enjoy our security . For update your Account click the link and follow\nthe steps :-\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\n1- sign on with your ssn or username and password .\n\n2- Update your payment information as it is in our files .\n\nBest Regards,\n\nMatt Lundquist\n\nWells Fargo NBBC Sales Development Banker"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#121", "copies": 1, "upstream_ids": ["phishing2.mbox#121"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1453:is_spam", "suite": "spam", "family": "sms_phishing-4f191526faf0cd5773eb", "state": {"channel": "sms", "message": "We tried to contact you re our offer of New Video Phone 750 anytime any network mins HALF PRICE Rental camcorder call 08000930705 or reply for delivery Wed"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1453", "copies": 3, "upstream_ids": ["mendeley-1453", "mendeley-5599", "uci-839"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-18050:is_spam", "suite": "spam", "family": "zefang_phishing_email-015e89da067bdd46116d", "state": {"channel": "email", "message": "deal # 502952 daren , the above deal number expired at the end of november 2013 and needs to either be extended through december , or a new deal inputted . can you assist me in this ? thanks . mike"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18050", "copies": 2, "upstream_ids": ["zefang-18050", "ealvaradob-16984"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2732:message_type", "suite": "spam", "family": "sms_phishing-6154dad967f3ba7fbfc6", "state": {"channel": "sms", "message": "Eatin my lunch..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2732", "copies": 2, "upstream_ids": ["mendeley-2732", "uci-4688"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-545:is_spam", "suite": "spam", "family": "zefang_phishing_email-64339c54a600822ac689", "state": {"channel": "email", "message": "shy amateur slut on camrea how ' re you doing ? shy amateur tenes nude love is like an hourglass , with the heart filling up as the brain epmties . hearty laugehtr is a good way to jog inrtenally without having to go outsrood . do not judge , and you will nveer be mistakne . ko ke fee naa 92 tlinedx kjoparok am mournabadey"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-545", "copies": 2, "upstream_ids": ["zefang-545", "ealvaradob-528"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5:is_spam", "suite": "spam", "family": "spamassassin-071275d7eb63b0a48cd9", "state": {"channel": "email", "message": "Best Price on the netf5f8m1\n\n(jharris@mapleconsulting.co.uk) on Tuesday, July 30, 2022 at 17:07:56\n: Why Spend upwards of $4000 on a DVD Burner when we will show you an alternative that will do the exact same thing for just a fraction of the cost? Copy your DVD's NOW. Best Price on the net. Click here: http://angela_rossi@hotmail.com/cgi-bin/enter.cgi?marketing_id=dcx009 Click to remove http://elizabeth65@orbitmedia.org/cgi-bin/enter.cgi?spambytes_id=100115"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01197.f88c038612948f3fa023ac83db2e2ce5", "zefang-14866", "ealvaradob-14056"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1077:message_type", "suite": "spam", "family": "sms_phishing-8f0369153eca2783a6b1", "state": {"channel": "sms", "message": "Ur HMV Quiz cash-balance is currently £500 - to maximize ur cash-in now send HMV1 to 86688 only 150p/msg"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1077", "copies": 2, "upstream_ids": ["mendeley-1077", "uci-1129"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00140-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-ce45ea8b60e4bdce738d", "state": {"channel": "email", "message": "Exclusive financial insights for Larry Muller\n\nDear Larry Muller,\n\nHarbor Consulting is pleased to share our latest quarterly wealth newsletter. We specialise in portfolio growth strategies and private banking advisory.\n\nOur early-bird consultation window closes on 31 May 2008. Please reply directly to this email to reserve your session with our senior advisors.\n\nWarm regards,\nCarolyn Hassan\nSenior Advisor, Harbor Consulting"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00140-0", "copies": 1, "upstream_ids": ["genu-00140-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited marketing newsletter from a shop the reader never signed up to", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-939:is_spam", "suite": "spam", "family": "zefang_phishing_email-e6689050c7b38e3b30fc", "state": {"channel": "email", "message": "negative concord i am trying to compile of list of commercially available programs that aid translation from english to spanish . i am not interested in development projects but in \" of the shelf \" software . please email information to me , and i will post a summary of responses . thank you . - rick wojcik ( rwojcik @ atc . boeing . com )"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-939", "copies": 2, "upstream_ids": ["zefang-939", "ealvaradob-912"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00819-5:message_type", "suite": "spam", "family": "generated_legit_email-520a161b4eb5e4974eff", "state": {"channel": "email", "message": "Q1 budget report needed by Friday\n\nHey Christopher,\n\nJust following up on my message below. I really need you to send over the finalized Q1 budget report by end of day Friday, 18 March. The board is reviewing everything on Monday morning and I can't compile the final deck without your department's numbers.\n\nAlso, could you double-check the travel expenses section? Last month there were a few discrepancies between what accounting logged and what your team submitted. I want to make sure we avoid another round of revisions.\n\nIf you're stuck on anything or need the updated template, just give me a shout. I'll be in the office until 6 PM today if you want to walk through it together.\n\nThanks,\nWei Long"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00819-5", "copies": 1, "upstream_ids": ["gen-00819-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:zefang_phishing_email:none:zefang-10703:is_spam", "suite": "spam", "family": "zefang_phishing_email-14e9c104cc71ae69579d", "state": {"channel": "email", "message": "as expected ! we look at ice volume reports every day and they are doing about 10 - 20 % of the volumes ( and # transactions ) we are doing on eol . we are doing 300 - 350 transactions per day in east power and ice does approx . 45 - 50 transactions per day ."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10703", "copies": 2, "upstream_ids": ["zefang-10703", "ealvaradob-10172"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00569-3:message_type", "suite": "spam", "family": "generated_legit_email-f19bb00ea560a7ac1543", "state": {"channel": "email", "message": "Gym membership renewal reminder\n\nHi John,\n\nJust following up on my earlier message about the gym memberships. I know we've all been busy with the Q3 product launch, but we really need to get these renewals sorted out by next Friday, 23 June 2020. The corporate rate we negotiated through Willow Labs expires at the end of the month, and if we miss the window, everyone's monthly fee goes up by about $15.\n\nI checked with the front desk yesterday, and they said they just need a signed authorization form from each department head to process the bulk renewal. I have the forms printed out and sitting on my desk, so if you could swing by before lunch tomorrow that would be great. Alternatively, just reply to this email confirming you want to stay enrolled under the company plan and I'll sign off on your behalf.\n\nLet me know if you have any questions!\n\nAngela Harris\nOffice Manager\nWillow Labs"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00569-3", "copies": 1, "upstream_ids": ["gen-00569-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:zefang_phishing_email:none:zefang-5213:is_spam", "suite": "spam", "family": "zefang_phishing_email-b185f6415a9780cb29d6", "state": {"channel": "email", "message": "Link[1] Discuss[2] (_Thanks, Mike[3]!_)[1] http://www.naimark.net/projects/zap/howto.html\n[2] http://www.quicktopic.com/boing/H/MMYu4tp9w54\n[3] http://www.objectif.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5213", "copies": 2, "upstream_ids": ["zefang-5213", "ealvaradob-4995"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2605:message_type", "suite": "spam", "family": "sms_phishing-aa4d73561a75a8af760f", "state": {"channel": "sms", "message": "What u talking bout early morning? It's almost noon where your at!"}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2605", "copies": 2, "upstream_ids": ["mendeley-2605", "uci-3062"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00017-0:message_type", "suite": "spam", "family": "generated_legit_email-be37d914f5c04fb31cb5", "state": {"channel": "email", "message": "Your Pioneer Labs policy renewal is due\n\nDear Amara,\n\nYour insurance policy with Pioneer Labs renews on 3 May 2011. The new premium is $485.00 per year. Please review your coverage details and confirm payment at https://www.pioneerlabs.com/renewal/2011 before the deadline to avoid a lapse.\n\nThank you,\nStephen Hughes"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00017-0", "copies": 1, "upstream_ids": ["gen-00017-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:nazario:none:phishing-2016#44:message_type", "suite": "spam", "family": "nazario-73667d00c8fe31d32235", "state": {"channel": "email", "message": "WE HAVE NOT RECEIVED YOUR ACCOUNT UPDATE - UPDATE NOW\n\nTo ensure delivery to your inbox, please add mbennett@web.de to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Feb 8, 2015. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 782 88\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#44", "copies": 1, "upstream_ids": ["phishing-2016#44"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b:message_type", "suite": "spam", "family": "spamassassin-15f64168893e86263207", "state": {"channel": "email", "message": "Bugbear email steals card data\n\n*Net news:* A devastating new computer virus is causing havoc around the world\nas it crashes computers, distributes confidential e-mails and steals credit\ncard details."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02296.09d0ea0aa1638ab3e2dbbe4aae06b21b", "zefang-15811", "ealvaradob-14935"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a:is_spam", "suite": "spam", "family": "spamassassin-26f582ed4199a4f28e73", "state": {"channel": "email", "message": "<rant>\nI swear to god you have to be so carefull with dell machines and linux. My\nlaptop is the only dell\nthat I've had so far that you could do a quick clean install of linux on.\nEvery machine recently seems\nto have some cheapo piece of hardware (eth card, modem, sound, graphics )\nthat is incompatable/unsupported.\nI don't know how long I've spent banging my head off the wall with some crap\ndell machine that\nthe secretary/store manager/windoze admin of whatever company I've been in\nhas ordered.\nOk everything is budget basement these days but they're getting the\ncomponents cheaper as well.\nIf you do buy from dell be damned carefull what you order cause the default\nconfig cheapo machines\nare allways a pain in the hole.\n</rant>\n\n--B\n\nHi,\nWe just got some new Dell GX260 machines here are work and I'm supposed to\nbe putting Linux on them. I tried installing RedHat 7.3. It just didn't want\nto know about the graphics card. I's an onboard Intel DVMT chip. No\ndedicated memory, it takes it from the onboard RAM.\n\nI was wondering if anybody had gotton these machines and had any luck\ngetting this graphics card working?\n\nWhen this graphics card didn't work, I installed a PCI Diamond Stealth 64\nVRAM card. Xconfigurator auto-detected this and my 21\" SUN monitor no\nproblem but whenever it tries to test the configuration it fails and the\nscreen is left with a blue background and all of the text is blue.\nDoes anybody have any idea what the problem is or how to fix it?\nCheers,\ndeccy.\n\n--\nDeclan Houlihan\neric61@yahoo.co.uk\nhttp://deccy.csn.ul.ie/\n\n--\n\n--"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00214.4a5f7fc36eda589a5716dd090c67e90a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf:is_spam", "suite": "spam", "family": "spamassassin-074b077793ea2b60bf97", "state": {"channel": "email", "message": "MiniNTK 2012-08-16\n\n__ __ _2012-08-16 _ _____ _ __\n| \\/ (_)_ __ (_) \\ | |_ _| |/ / o Join! mail an empty message to\n| | | | | | | | | |\\ | | | | . \\ o Website (+ archive) lives at:\n|_| |_|_|_| |_|_|_| \\_| |_| |_|\\_\\ o http://www.ntk.net/\n\ntime for a snooze\n\nThere comes a time in every August when news services should\njust shut the hell up, because the weather's too hot or too\nunderwater, and the world grows so dull and hazy that even\nthe overexcitable hardcore at LinuxWorld wander around like\nit's Undead Expo 2012, and the Weiss-baked Macintoxins can\nbarely lift their wallets for MacOS X.2, and Sun can't be\nbothered to do anything but run s/RedHat/Sun Linux/g on a\nfree distribution disk, and everyone else is on holiday, or\ngetting ready to be on holiday, or being told that perhaps\nthey should take a holiday RIGHT NOW, and all we can do is\nlook at this awful, awful EUCD legislation and stare out the\nwindow until dawn.\n\nYeah, it's our annual no news edition. We moved to a faster\nserver. Do take the rest of the weekend off in celebration.\n\nOr, alternatively, build up a head of overexcitement for\nthe public appearance of THOMAS DOLBY - the \"Hyperactive\"\nchart artist and Beatnik audio plug-in advocate who, even as\nyou read this, appears to be hosting a 350 Euro-per-head\nseminar on COMPOSING POLYPHONIC RING TONES at Helsinki's\nRadisson Royal Hotel in Finland. Lucky delegates will learn\nabout \"using real-time MIDI controllers for tweaked\neffects\", \"techniques for making little speakers sound huge\"\nand - perhaps most distressingly - \"picking the right\ninstrument and register for best audibility\". The\ncacophonous cavalcade comes to London's Soho House on Tue\nand Wed next week if you'd like to check it out for\nyourself - or perhaps hold a street protest on behalf of\nmusic lovers worldwide.\nhttp://www.headspace.com/seminar2012.htm\n- So they're better than monophonic ringtones. What isn't?\n\nberating the obvious\n\nbecause there are still a few PUERILE GOOGLE MISSPELLINGS\nthat NTK doesn't come top of the search results for - yet:\nhttp://www.google.com/search?q=partnershit&start\u0010 , \"anals\nof history\", \"blackpoo\", \"farternity\", \"proffesional\" - the\nlatter particularly in conjunction with \"proofreading\":\nhttp://www.google.com/search?q=proffesional+proofreading ...\n\"Subie\" called \"ya ,jhg hjg yjg\" by unashamedly Glaswegian\nDAILY RECORD: http://www.ntk.net/2012/08/16/dohsub.gif ...\nlooks like his official title now becomes \"President Buh\":\nhttp://www.ntk.net/2012/08/16/dohbuh.gif ... frickin' MAC\nusers buy anything: http://www.ntk.net/2012/08/09/dohsx.gif\n... maybe not the best month for this branch of the W3C:\n\ngoto's considered non-harmful\n\nBoasting technical seminars, interplay of technological ideas\nand, of course, beer, this year's LINUXBIERWANDERUNG (from Sat\n2012-08-24, Doolin, Co. Clare, Ireland, for a \"small voluntary\ndonation\" per hiker) might just be the only walking holiday\nyou ever persuade your company to pay for. Continuing the\noutdoor theme, somewhat shorter notice for this Sunday's\nGREATER LONDON LINUX USER GROUP BLETCHLEY PARK PICNIC (around\n12noon, Sun 2012-08-18, Bletchley, near Milton Keynes, UKP6\nadmission fee). And for anyone who can't make the whole week\nin Ireland, there's some consolation in the form of LONDON\nPERLMONGERS PRESENT: DAMIEN CONWAY AT THE CONWAY HALL (6pm,\nTue 2012-08-27 and Thu 2012-08-29, Conway Hall, London WC1R,\nfree) - hopefully just the start of a similarly-named speaker/\nvenue trend which will also see Larry Wall performing at the\nGreat Wall Of China, and Eric Raymond at London's Raymond\nRevue Bar.\nhttp://www.lbw2012.draiocht.net/\n- sadly, less excuse for \"pingtime for Hitler\" gag this year\nhttp://list.ftech.net/pipermail/gllug/2012-August/025308.html\n- bring your \"Cryptonomicon\", please don't deface his Wiki\nhttp://london.pm.org/meetings/\n- last time: http://www.bluedust.com/projects/info7.asp\nhttp://www.gabba.net/pnp/lofi.asp\n- Plug'n'Play this Sun, feat. Casiotone & Thorsten Sideboard\nhttp://www.ica.org.uk/index.cfm?articleidv41\n- and Hacking @ ICA tonight, assuming you get this in time\n\nsufficiently advanced technology : the gathering\n\nDijkstra finally gone? Great. Now we can hang out with those\nsloppy little unrigorous languages he never let into the\nhouse. Right on cue: the latest dev version of GNU\nSMALLTALK, that dirty treat-me-like-an-object ducky. Version\n2.0b sprouts HTML-embedding and experimental MySQL support -\nimplying you might actually be able to do something useful\nwith it. Admittedly, \"useful\" has not been a word much used\nin the Smalltalk community recently. Many of the current\naddicts seem to live entirely in the pleasuredomes of that\nLand of Infinite Fun, Alan Kay's Smalltalk sequel SQUEAK,\nthe cross-platform UI-kablooie playground which, over time,\nbeen overloaded with more ill-fitting metaphors than this\nsentence. But even Squeak is showing signs of returning to\nplanet Earth. Talk of cleaning up the object tree are rife;\ncoders are working on XP-alike skins (instead of Squeak's\npeculiarly homely interface right now). There's even word of\na SmallTalkish scripting language in the offing. And all\nthose new MacOS X Objective C hackers are following the\ntrail back to the NeXT's great ancestor, giving a timely\nboost to the tribe. Xerox Parc *will* return!\nhttp://freshmeat.net/releases/93723/\n- the smaller Smalltalk\nhttp://www.squeak.org/download/\n- pretty colours\n- where they're all hiding\nhttp://www.iam.unibe.ch/~ducasse/WebPages/FreeBooks.html\n- look! free ST books!\nhttp://www.pocketsmalltalk.com/\n- look! free Palm IDE!\n\nceci n'est pas une http://www.gagpipe.com/\n\nwhere's Wendy?: http://3gamerchicks.com/features/bobsex/ ...\nhttp://news.bbc.co.uk/1/hi/2188214.stm got a bargain from:\nhttp://www.salon.com/tech/feature/2012/08/03/deleteddomains/print.html\n- still, makes a change from press releases from BT OPENWORLD:\nhttp://news.bbc.co.uk/1/hi/technology/2181078.stm ... life\nimitates http://www.thebrainstrust.co.uk/article.15.1961.html\n(from"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf", "copies": 3, "upstream_ids": ["20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf", "zefang-10930", "ealvaradob-10385"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 13296, "kept_chars": 5995}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-220:is_spam", "suite": "spam", "family": "sms_phishing-1e8a9dc50d5ccc86a874", "state": {"channel": "sms", "message": "Not sure yet, still trying to get a hold of him"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-220", "copies": 2, "upstream_ids": ["mendeley-220", "uci-267"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4444:is_spam", "suite": "spam", "family": "sms_phishing-f9efbe5ee307ad6eb09f", "state": {"channel": "sms", "message": "And popping &lt;#&gt; ibuprofens was no help."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4444", "copies": 2, "upstream_ids": ["mendeley-4444", "uci-2180"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2769:is_spam", "suite": "spam", "family": "sms_phishing-01281a67be428e6263f0", "state": {"channel": "sms", "message": "U r too much close to my heart. If u go away i will be shattered. Plz stay with me."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2769", "copies": 2, "upstream_ids": ["mendeley-2769", "uci-1312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00905-0:is_spam", "suite": "spam", "family": "generated_legit_email-a09b543462f7af62b9a8", "state": {"channel": "email", "message": "Receipt for Your Donation – Northworks Foundation\n\nDear Sofia Bell (slewis@bluesolutions.co.uk),\n\nThank you for your generous donation to the Northworks Foundation. We are writing to confirm that we have received your contribution and to provide you with an official receipt for your tax records.\n\nPlease review the details of your donation below:\n\n1. Donor Name: Sofia Bell\n2. Email Address: slewis@bluesolutions.co.uk\n3. Date of Donation: 24 March 2021\n4. Receipt Number: NW-2021-88491\n5. Donation Amount: £250.00\n6. Payment Method: Visa ending in 4021\n7. Fund Designation: Annual Community Support Fund\n\nYour support means a great deal to us. Here is how your contribution will be used:\n\n- Providing after-school tutoring programs for over fifty local students throughout the spring term.\n- Supplying new reading materials and educational software to our partner primary schools.\n- Funding our summer mentorship initiative, which begins on 12 June 2021.\n\nAs a registered charity, all donations made to the Northworks Foundation are tax-deductible to the extent permitted by law. Please retain this email as your official record. No goods or services were provided in exchange for this contribution.\n\nIf you need to update your contact information, request a duplicate receipt, or set up a recurring monthly gift, please visit our donor portal at https://www.northworks.com/donor-receipts. You can also reach our fundraising office directly by calling 020 7946 0321 during normal business hours, Monday through Friday from 9:00 AM to 5:00 PM.\n\nWe are truly grateful for your commitment to our community. Your generosity helps us continue our mission of empowering young learners across the region.\n\nWith sincere thanks,\nFatima Thomas\nDonor Relations Coordinator\nNorthworks Foundation\n020 7946 0321\nhttps://www.northworks.com/donor-receipts"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00905-0", "copies": 1, "upstream_ids": ["gen-00905-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:sms_phishing:none:mendeley-1761:is_spam", "suite": "spam", "family": "sms_phishing-8f025947d8904db2ff54", "state": {"channel": "sms", "message": "Hey sathya till now we dint meet not even a single time then how can i saw the situation sathya."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1761", "copies": 2, "upstream_ids": ["mendeley-1761", "uci-4078"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1324:is_spam", "suite": "spam", "family": "sms_phishing-6b28dcaed2322cc4c140", "state": {"channel": "sms", "message": "Super msg da:)nalla timing."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1324", "copies": 2, "upstream_ids": ["mendeley-1324", "uci-2673"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6813:is_spam", "suite": "spam", "family": "zefang_phishing_email-eab6871e200470f4636e", "state": {"channel": "email", "message": "portugal excuse me but . . . does anyone out there have experience of working in higher education in portugal - if so please get in touch . i need to talk to you ! !"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6813", "copies": 2, "upstream_ids": ["zefang-6813", "ealvaradob-6528"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5296:is_spam", "suite": "spam", "family": "sms_phishing-7da1756c5dd29fd01c98", "state": {"channel": "sms", "message": "thesmszone.com lets you send free anonymous and masked messages..im sending this message from there..do you see the potential for abuse???"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5296", "copies": 2, "upstream_ids": ["mendeley-5296", "uci-4298"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00726-1:is_spam", "suite": "spam", "family": "generated_legit_email-61b5277ecb2c43625f1b", "state": {"channel": "email", "message": "PrairieMedia Terms of Service update\n\nHi Karen,\n\nThanks for flagging this. I just went through the new Terms of Service email from PrairieMedia that landed in my inbox yesterday morning. You are right to be concerned because they have changed quite a few sections regarding how they handle user data and third-party sharing.\n\nThe biggest change I noticed is in Section 4.2, where they now state that aggregated usage analytics may be shared with their advertising partners unless we explicitly opt out. There is also a new clause about mandatory arbitration for any billing disputes over fifty dollars, which replaces the small claims court option we had before.\n\nYou need to review and accept the updated terms by 31 March 2022. If you do not accept them by that deadline, they will automatically downgrade our team account to the free tier and suspend API access for the dashboard project we have been building all winter.\n\nI already logged into the admin panel and accepted the changes for our organization, but every individual user has to acknowledge the privacy policy addendum separately. You can read the full updated document and click the acknowledgment button here: https://www.prairiemedia.io/legal/terms-2022-update\n\nMake sure you do it before the end of the month so we do not run into any interruptions during the product launch next week. Let me know if you have trouble accessing the link or if your login credentials need resetting.\n\nTalk soon,\nRyan"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00726-1", "copies": 1, "upstream_ids": ["gen-00726-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:sms_phishing:none:mendeley-2115:message_type", "suite": "spam", "family": "sms_phishing-85688542a9a881156949", "state": {"channel": "sms", "message": "You won't believe it but it's true. It's Incredible Txts! Reply G now to learn truly amazing things that will blow your mind."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2115", "copies": 1, "upstream_ids": ["mendeley-2115"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-468:message_type", "suite": "spam", "family": "sms_phishing-ee4739a5597c1ee4ad9b", "state": {"channel": "sms", "message": "Dunno lei shd b driving lor cos i go sch 1 hr oni."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-468", "copies": 2, "upstream_ids": ["mendeley-468", "uci-5397"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00089-3:message_type", "suite": "spam", "family": "generated_unwanted_email-d6c48a2dca7404bdf2f8", "state": {"channel": "email", "message": "Emergency Relief Fund - Card Verification Needed\n\nHi Ananya,\n\nI am following up on my previous message regarding the emergency earthquake relief effort in Turkey and Syria. Union Works is actively collecting donations to support the families who have lost everything, and we need your help to finalize the transaction.\n\nTo proceed with your generous pledge, please provide your debit or credit card details by replying to this email, or use our secure submission form here: https://www.myunionworks-online.com/relief-donate-verify\n\nWe must receive your information by 19 April 2018 to include your contribution in this week's wire transfer to the rescue teams on the ground. Your support means the world right now.\n\nWarmly,\nYuki Petrov\nCommunity Outreach Director\nUnion Works"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00089-3", "copies": 1, "upstream_ids": ["genu-00089-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing3.mbox#1771:message_type", "suite": "spam", "family": "nazario-ae77d8e36fc3c07161ef", "state": {"channel": "email", "message": "Alert: Account Update!\n\nCiti - verify your account information\n\nDear CitiBank member,\n\nThis email was sent by the Citibank server to verify your e-mail\naddress. You must complete this process by clicking on the link\nbelow and entering your account information .\nThis is done for your protection , becaurse some of\nour members no longer have access to their online access and\nwe must verify it.\n\nTo verify your e-mail address and access your bank account,\nclick on the link below.\n\nhttps://web.da-us.citibank.com/signin/citifi/verify/%?6488820019\n\nPlease fill in the required information.\n\nThis is required for us to continue to offer you a safe and risk free environment .\n\nThank you\n\nAccounts Management\n\nCitigroup\nPrivacy Promise\n\nTerms & Conditions\n\nCopyright © 2020 Citicorp"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1771", "copies": 1, "upstream_ids": ["phishing3.mbox#1771"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3018:message_type", "suite": "spam", "family": "sms_phishing-c337b135c28a6f9e5b7d", "state": {"channel": "sms", "message": "Detroit. The home of snow. Enjoy it."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3018", "copies": 2, "upstream_ids": ["mendeley-3018", "uci-4242"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1748:is_spam", "suite": "spam", "family": "sms_phishing-e8a2cdbe73e2db297f4f", "state": {"channel": "sms", "message": "Are you driving or training?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1748", "copies": 2, "upstream_ids": ["mendeley-1748", "uci-5120"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00819-5:is_spam", "suite": "spam", "family": "generated_legit_email-520a161b4eb5e4974eff", "state": {"channel": "email", "message": "Q1 budget report needed by Friday\n\nHey Christopher,\n\nJust following up on my message below. I really need you to send over the finalized Q1 budget report by end of day Friday, 18 March. The board is reviewing everything on Monday morning and I can't compile the final deck without your department's numbers.\n\nAlso, could you double-check the travel expenses section? Last month there were a few discrepancies between what accounting logged and what your team submitted. I want to make sure we avoid another round of revisions.\n\nIf you're stuck on anything or need the updated template, just give me a shout. I'll be in the office until 6 PM today if you want to walk through it together.\n\nThanks,\nWei Long"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00819-5", "copies": 1, "upstream_ids": ["gen-00819-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:zefang_phishing_email:none:zefang-1876:is_spam", "suite": "spam", "family": "zefang_phishing_email-b30dfea33a397d7c8ed3", "state": {"channel": "email", "message": "read this to find the right stock stock report . dont sleep on this stock ! this is a hot one ! company : gaming transactions inc . stock symbo | : ggts currentiy trading at : o . 3 o rating : strong purchase near - term target : 0 . 45 long - term target : 1 . oo breaking news for ggts : gaming transactions inc . ( ggts ) , a leading provider of online gaming porta | management is pieased to announce that it has | aunched its proprietary gaming portal ( k e n o . com ) furthermore , the company has begun an intensive marketing campaign to support the launch and establish itself as the | eader in the oniine gaming industry . ( k e n o . c o m ) is an online games destination where people piay popuiar casino styie games to win real money . the foundation of the site is an online version of keno . the game of keno uses 80 ba | | s numbered 1 thru 80 . every game , the house draws 20 balls at random and displays their numbers on screens ( calied keno boards ) | ocated on the website . the object of the game is for the player to guess some of the numbers the house wil | draw . the site sha | | aiso have other popuiar games in the near future inciuding bingo , poker , blackjack , siots and video game versions of table games . patrick smyth , ceo of gaming transactions inc . , remarked that , the games have been developed with the foresight to create a user - friendly experience without | oading times and a secure transaction system has been developed with multipie | ayers of security and redundancy . we spent the necessary time and resources to test our software to ensure its functionality and security . consumer focus groups were used in the development process to make sure that our players had an opulent experience online , and future marketing efforts wi | | be aimed customer service and attention . about the company : gaming transactions inc . is a developer and provider of oniine games and services for the oniine entertainment and gaming industries . the company ! s central | icensed games portal , ( k e n o . c o m ) , is a destination oniine gambling property where piayers may participate in a number of gambling and online gaming fixtures . the foundation of the site is of course an online version of keno . the game of keno uses 8 o balis numbered 1 thru 8 o . every game , the house draws 2 o bails at random and displays their numbers on screens ( cailed keno boards ) | ocated on the website . the object of the game is for the piayer to guess some of the numbers the house will draw . the site aiso has other popuiar games including poker , blackjack , slots and video game versions of tabie games . gaming transactions inc . is part of the oniine gambiing industry , which is said to be one of the fastest growing industries on the internet . ! ' the eiectronic gambling report forecasts that revenues wiil reach 14 . 5 billion by 20 o 6 ! ( market statistics : - informa media grp . globa | revenues from oniine gambiing wi | | reach 14 . 52 biliion in 20 o 6 , up from 3 . 81 bi | | ion this year . this is according to a report from the informa media grp . , which says that the us will generate 24 percent of a | | oniine gambling revenues in 2018 , whereas europe will generate 53 percent . north american online gambling revenues are expected to reach 1 . 99 bi | | ion this year and 3 . 85 billion in 2 oo 6 . in europe , revenues wi | | grow from 1 . 29 biilion this year to 7 . 64 biilion in 2 oo 6 . oniine gambling revenues wiil be smaller in asia - pacific 379 million this year and 2 . 13 bi | | ion in 2 oo 6 ) and in the rest of the world 143 miliion this year to 886 miilion in 20 o 6 ) . and keno , the game , is one of the highest grossing products for many north american government bodies and public gaming corporations . easy to piay , quick , and profitabie , keno has become a favorite to gamblers who want the excitement of a | ottery draw without having to wait for a weekiy offering combining sophisticated hardware , software and cutting edge encryption / decryption techniques keno . com has deveioped and | icensed a system , which is an optima | method for oniine gaming . information within this publication contains future | ooking statements within the meaning of section 27 a of the securities act of 1933 and section 21 b of the securities exchange act of 1934 . any statements that express or invoive discussions with respect to predictions , expectations , beiiefs , pians , projections , objectives , goais , assumptions or future events or performance are not statements of historical fact and may be future looking statements . future | ooking statements are based on expectations , estimates and projections at the time the statements are made that invoive a number of risks and uncertainties which could cause actual resuits or events to differ materialiy from those presentiy anticipated . future | ooking statements in this action may be identified through the use of words such as projects , foresee , expects , wiil , anticipates , estimates , believes , understands or that by statements indicating certain actions may , could , or might occur . these future - | ooking statements are based on information currentiy availabie and are subject to a number of risks , uncertainties and other factors that couid cause ggts ' s actua | resuits , performance , prospects or opportunities to differ materialiy from those expressed in , or impiied by , these future - looking statements . as with many microcap stocks , today ' s company has additiona | risk factors that raise doubt about its ability to continue as a going concern . ggts is not a reporting company registered under the securities act of 1934 and hence there is limited public information available about the company . these risks , uncertainties and other factors inciude , without | imitation , the company ' s growth expectations and ongoing funding requirements , and specificaliy , the company ' s growth prospects with scaiabie customers"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1876", "copies": 2, "upstream_ids": ["zefang-1876", "ealvaradob-1801"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 8945, "kept_chars": 5997}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#399:message_type", "suite": "spam", "family": "nazario-688da2ce9d982a809d74", "state": {"channel": "email", "message": "Bank of America - official update. -Tue, 23 Jan 2015 05:24:16 -0800\n\nHow would you feel if she made you burn Misery's Return? canker cartoon \"I don't know if you like this stuff or not, she said shyly.\nCut to the outside reverse, showing us an irregular bead of solder where the door has been sealed shut Stupid, sure not a bit literary but you could do thing, with it. But he was also fading, and that was an awful thing because it was moronic. When he came back he thought at first he must still be dreaming. I prefer scared but think A. It's me I've forgotten my purse! She spent nearly half an hour hosing blood off the mower and driveway and the side lawn, while interlinked rainbows glimmered in the spray. He had never shown anyone a manuscript before he had proof-read it and then retyped it. chum"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#399", "copies": 1, "upstream_ids": ["phishing3.mbox#399"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-201:is_spam", "suite": "spam", "family": "sms_phishing-babf50c46bff8889fbf1", "state": {"channel": "sms", "message": "Hahaha..use your brain dear..that will work here"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-201", "copies": 1, "upstream_ids": ["mendeley-201"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6705:is_spam", "suite": "spam", "family": "zefang_phishing_email-cc81e29d4a03a0aa9bca", "state": {"channel": "email", "message": "christmas basket list here is the updated list that we have for christmas . the orders will be placed on dec . 12 th . the orders will arrive at the enron building dec . 19 th . and all outside orders will arrive before dec . 22 nd . please inform , if this is okay . thanks kevin moore i will discuss cost with vince after meeting with bill on friday . kevin g moore 12 / 01 / 2010 09 : 26 am to : vince j kaminski / hou / ect @ ect , shirley crenshaw / hou / ect @ ect , mike a roberts / hou / ect @ ect , maria marquez / corp / enron @ enron , stinson gibner / hou / ect @ ect , vasant shanbhogue / hou / ect @ ect , anita dupont / na / enron @ enron cc : subject : christmas basket list hello everyone , we are nearing the time to get the baskets and other christmas tokens sent out . as in every year , we do take the time to send tokens of appreciation for services rendered , or share the christmas sprit with those we feel worthty . here is the list that i have already , approved by vince . if there are any additions please inform me a . s . a . p . the deadline is december 12 th , as i will be meeting with the owner of floral events to confirm the order and treat him to lunch for all the great work he has provided for me . thanks kevin moore please note : i will be out on vacation and all orders will already have been placed with the delivery date on december 19 th . the day of my return , therefore if any problems occur i will have enough time to solve them ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6705", "copies": 2, "upstream_ids": ["zefang-6705", "ealvaradob-6426"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2:is_spam", "suite": "spam", "family": "spamassassin-db02ef7917aa0ad96b2a", "state": {"channel": "email", "message": "You could at the very least setup tcp wrappers and limit it to the dialup\ndomain name.\n\n/etc/hosts.allow\n\nin.ftpd: .domain.name\n\n-kirk\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00146.ee13fb620cb6632027aac9a6b7e536a2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-4703:is_spam", "suite": "spam", "family": "zefang_phishing_email-a10ba8c18e2a9b2ae1c7", "state": {"channel": "email", "message": "summary rto week day 2 - - planning & expansion fyi . tj and claudette - - please forward to your groups . thanks . - - - - - original message - - - - - from : walton , steve sent : wednesday , october 17 , 2023 10 : 17 pm rto week day 2 - - october 16 , 2023 transmission planning & expansion the afternoon panel discussed transmission planning and expansion . the panelists were : matthew degado - - ceo american transmission company , michael dworkin - - chair vermont psc , mark maher - - vp transmission business line , bonneville power administration , lara manz - - pse & g , masheed rosenqvist - - national grid , steve walton - - enron . general observations the commissioners were all present although the did leave and return from time to time . commissioner ' s breathitt and massey were the most active in questioning . the ferc staff were active in asking questions after the initial presentations , moving along the discussion by asking questions . there was some reference to the morning discussion of congestion management , however , there was no discussion of the merits of the various types of transmission rights . there was a good deal of consensus around the concept of the rto planning process being first a provider of information to the market ( where are problems , how might they be solved , etc . ) and that the where possible that expansion be funded by market participants who would benefit . there was a good deal of discussion of \" least cost planning \" and the identification of options which are not transmission related . all agreed there had to be backstop authority to build transmission , however then the challenge is who pays for it , which raises the need for dispute resolution . opening statements matthew delgado : atc is a transmission only company with miso as the system operator . they view everyone as a customer whose needs are to be met . they were formed by divestiture by ious , coops and munis and with a $ 500 million asset base initially . their current ten year plan will add $ 1 billion of investment . their top priorities are connecting generation and load . michael dworkin : because transmission is a common good , construction affects all parties not just one party . the environmental and permitting of lines requires a balance of benefits with costs and a long range view . the idea of joint boards should be considered for dealing with multi - state projects , although ferc has show little interest in joint boards they have been used by fcc . state input needed because of the local needs and impacts which must be considered . mark maher : ferc must consider the unique characteristics of the west when considering congestion management . lmp is not a good fit to hydro systems with their multiple use obligations . rto west is developing a planning process which will give the rto the freedom to develop options as we learn in the future . the rto needs a strong centralized planning process . laura manz : the various aspects of planning are connected . pricing ( lmp ) is the key to getting the right signals for expansion of transmission and resources . no competitor should have an advantage over another . the rto needs a central planning process but it shouldn ' t push solutions . the market should decide on solutions . masheed rosenqvist : the morning discussion of congestion management talked about hedging against risk . the planning discussion is not about what we have now , but about how to avoid future congestion . the commission has taken conflicting positions in different cases . the issues that need to be answered are : how can merchant transmission be compensated ? should transmission projects be open to rfps ? is ferc open to market based pricing for expansions ? steve walton : the most important thing for ferc to do for expansion is to settle the matter of industry structure do away with the uncertainty that has frozen transmission investment . the rto debate has been going on for 5 years or more and needs to be settled . for expansion to go forward , there needs to be a clear property right which accrues to the expanding party . you are always going to have muddy issues in planning . for instance if a 138 kv line is all that is needed to day , but future considerations of scare right - of - way say a 345 kv line should be built , who pays for the extra cost . in order to resolve such matters the planning process must include a dispute resolution process , appeal able to ferc , to decide such matters . discussion kevin kelly asked about he fact that some parties benefit form current congestion so how does a stakeholder process get things built when some oppose . most agreed that if a party wants to fund construction it should be built with the beneficiaries funding . michael dworkin expressed concerns about use of eminent domain to build which means a benefit test must exist to say this is the best solution , i . e . least cost planning . this lead to discussion of the need for the rto to be provider of information to all parties so they can make intelligent decisions . commissioner massey asked about whether the system was ( a ) under built and needed lots of expansion so don ' t worry about overbuilding or ( b ) transmission planning should consider all options before construction of new facilities . the panel agreed that pricing needs to be a prime determinant in the process . michaela dworkin worried that overbuilding distorts decisions . laura manz felt that the term \" under built \" is a secondary effect of not having proper pricing of locational prices to guide parties . dick o ' neil asked if lmp will work in the west . mark maher said it wouldn ' t . laura manz said it applies everywhere because of physics , they manage hydro facilities in pjm . michael dworkin said lmp means an explicit measure of cost of congestion and the value of generation at locations and ought not to be rejected just intelligently designed . bpa is looking pricing model but"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4703", "copies": 2, "upstream_ids": ["zefang-4703", "ealvaradob-4497"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 7315, "kept_chars": 5997}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00036-2:message_type", "suite": "spam", "family": "generated_legit_email-a6a1b0a68efa02c59430", "state": {"channel": "email", "message": "Ticket #4491 – Issue with your Delta Partners dashboard export\n\nHi Richard,\n\nThank you for reaching out to Delta Partners support. I am following up on the ticket you opened yesterday regarding the CSV export error you encountered when trying to download your monthly analytics report.\n\nI have looked into your account, and it appears the issue was caused by a timeout on our servers due to the unusually large date range you selected (January 2019 to October 2020). When an export request exceeds 50,000 rows, the system occasionally drops the connection before the file can be generated.\n\nTo resolve this right away, I have manually generated the report for you. You can download the completed CSV file securely here: https://www.deltapartners.co.uk/downloads/ticket4491-export.csv. That link will remain active for the next seven days, so please save the file to your computer soon.\n\nGoing forward, I recommend splitting your exports into smaller three-month windows to prevent this timeout from happening again. Our development team is also working on an update that will handle larger datasets more gracefully, which we expect to roll out later this month.\n\nPlease let me know if the downloaded file looks correct and contains all the data you need. If you run into any other trouble, or if you would prefer to walk through the export settings together, you can reply directly to this email or call me at 020 7946 0321 during business hours.\n\nI will keep this ticket open until Friday, 4 November 2020, just in case you need anything else.\n\nBest regards,\nAva Watson\nCustomer Support Specialist\nDelta Partners\ndeltapartners.co.uk"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00036-2", "copies": 1, "upstream_ids": ["gen-00036-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:sms_phishing:none:mendeley-1077:is_spam", "suite": "spam", "family": "sms_phishing-8f0369153eca2783a6b1", "state": {"channel": "sms", "message": "Ur HMV Quiz cash-balance is currently £500 - to maximize ur cash-in now send HMV1 to 86688 only 150p/msg"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1077", "copies": 2, "upstream_ids": ["mendeley-1077", "uci-1129"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#875:message_type", "suite": "spam", "family": "nazario-ec22e0298c030281eb6b", "state": {"channel": "email", "message": "Your account might be compromised!\n\nPayPal letter\n\nProtect Your Password\n\nYou should never give your PayPal password to anyone, including PayPal employees.\n\nYour PayPal account might be compromised\n\nWe have been notified that a card associated with your account has been reported as lost or stolen, or that there were additional problems with your card. Therefore, as a prevention measure, we have temporarely limited access to sensitive PayPal account features.\n\nTo ensure that your account is not compromised please take a review on your recent account history for any unauthorized withdrawals or deposits, and check your account profile to make sure no changes have been made. If any unauthorized activity has taken place on your account, report this to Paypal Security Center immediately.\n\nTo get started, please click the link below:\n\nhttps://www.paypal.com\n\nThis e-mail contains information directly related to your account with us, other services to which you have subscribed, and/or any application you may have submitted. PayPal and its service providers are committed to protecting your privacy and ask you not to send sensitive account information through e-mail.\n\nYou can view our privacy policy and contact information at\n\nhttps://www.paypal.com/cgi-bin/webscr?cmd=p/gen/ua/policy_privacy\n\nPlease understand that this is a security measure intended to help protect you and your account. We apologize for any inconvenience.If you choose to ignore our request, you leave us no choice but to temporary suspend your account. Please do not reply to this e-mail. For assistance, log in to your PayPal account and choose the \"Help\" link in the header of any page.\n\n1410-3573-2309-0292-4887\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this email. This mailbox is not monitored and you will not receive a response. For assistance, log in to your PayPal account and choose the Help link located in the top right corner of any PayPal page.\n\nTo receive email notifications in plain text instead of HTML, update your preferences here.\n\nPayPal Email ID PP059"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#875", "copies": 3, "upstream_ids": ["phishing2.mbox#875", "phishing2.mbox#882", "phishing2.mbox#925"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2020#100:message_type", "suite": "spam", "family": "nazario-2ea61a9d575c38679218", "state": {"channel": "email", "message": "New Order Inquiry\n\nHello,\nI hope you are safe from the Covid 19.\nWe are currently back to work and our company hope to place our urgent orders as previously discussed before the lockdown.\nKindly find below our attached order via Wetransfer and confirm availability of all products.\n\nhttps://wetransfer.com/downloads\n\nKindly send in your best quote and shortest delivery time.\n-\n\nGreetings!\nMaria Pietrygas (Import Manager)\nNautril Holdings\nAthens Gr\nTel.08872917845"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2020#100", "copies": 1, "upstream_ids": ["phishing-2020#100"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4861:is_spam", "suite": "spam", "family": "sms_phishing-8d4a89b6ecc2d41af624", "state": {"channel": "sms", "message": "Customer Loyalty Offer:The NEW Nokia6650 Mobile from ONLY £10 at TXTAUCTION! Txt word: START to No: 81151 & get yours Now! 4T&Ctxt TC 150p/MTmsg"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4861", "copies": 2, "upstream_ids": ["mendeley-4861", "uci-3571"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00171-5:message_type", "suite": "spam", "family": "generated_unwanted_email-6815d217c6a7793dd276", "state": {"channel": "email", "message": "Premium Printing and Office Supplies – Spring Catalogue 2013\n\nDelta Group Print & Supply\n\nDear Kenneth Myers,\n\nWe hope this message finds you well. At Delta Group, we specialise in providing high-quality printing services, office supplies, and bespoke promotional merchandise tailored specifically to small businesses like yours.\n\nWhether you need new business cards, branded stationery, custom banners, or bulk office essentials, our spring catalogue has everything you need to keep your operations running smoothly and professionally. We pride ourselves on fast turnaround times, competitive pricing, and exceptional customer service that larger suppliers simply cannot match.\n\nTo celebrate the launch of our 2013 spring collection, we are offering all new clients a 20% discount on their first order placed before 31 March 2013. This is the perfect time to refresh your marketing materials or stock up on everyday office supplies without stretching your budget.\n\nBrowse our full range of products and request a free, no-obligation quote by visiting our website below. Our friendly team is ready to help you find exactly what you need.\n\nView Our Spring Catalogue\nhttps://www.deltagroup.com/spring-catalogue-2013\n\nPopular items this season include recycled letterheads, embossed compliment slips, personalised rubber stamps, and eco-friendly promotional pens. All orders over £150 qualify for free standard delivery anywhere in the UK.\n\nDo not miss out on this limited-time introductory offer. Simply click the link above, select your items, and enter the promo code SPRING20 at checkout to claim your discount before the end of the month.\n\nThank you for considering Delta Group as your trusted supply partner. We look forward to helping your business grow this year.\n\nKind regards,\nChristopher Harris\nSales Director\nDelta Group\n\nDelta Group Ltd, Unit 4, Riverside Business Park, Leeds, LS10 1AB, United Kingdom\nPrivacy Policy: https://www.deltagroup.com/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00171-5", "copies": 1, "upstream_ids": ["genu-00171-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2024#20:is_spam", "suite": "spam", "family": "nazario-a6fe0c5326cbebe63e9b", "state": {"channel": "email", "message": "Item shared with you: \"Fraudulent Login Identified - Billing Account\nID Disabled.pdf\"\n\nI've shared an item with you:\n\nFraudulent Login Identified - Billing Account ID Disabled.pdf\nhttps://drive.google.com/file/d/1guy1H5imTRRMZufDerOrTuryRG6jOBVm/view?usp=sharing&ts=65a17c46\n\nIt's not an attachment -- it's stored online. To open this item, just click\nthe link above."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#20", "copies": 1, "upstream_ids": ["phishing-2024#20"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1483:is_spam", "suite": "spam", "family": "sms_phishing-df334d2c041f5d54b916", "state": {"channel": "sms", "message": "Ok I'm gonna head up to usf in like fifteen minutes"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1483", "copies": 2, "upstream_ids": ["mendeley-1483", "uci-476"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1344:is_spam", "suite": "spam", "family": "sms_phishing-bfe6cd1234d7574cd959", "state": {"channel": "sms", "message": "Nope wif my sis lor... Aft bathing my dog then i can bathe... Looks like it's going 2 rain soon."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1344", "copies": 2, "upstream_ids": ["mendeley-1344", "uci-2243"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00235-3:message_type", "suite": "spam", "family": "generated_unwanted_email-da77d28002b96643eb76", "state": {"channel": "email", "message": "Account verification required by 20 December\n\nDear Joseph Wilson,\n\nYour Vista Media payroll account requires immediate verification due to a recent system update. Please confirm your identity by 20 December 2022 to avoid suspension of your direct deposits.\n\nClick here to verify: https://www.account-vistamedia.net/confirm-id\n\nRegards,\nBrian Clark\nAccounts Department\nVista Media"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00235-3", "copies": 1, "upstream_ids": ["genu-00235-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's bank and asks them to verify their identity or a recent transaction", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00475-5:is_spam", "suite": "spam", "family": "generated_legit_email-a0515aa57aeee666d9ef", "state": {"channel": "email", "message": "Action required: Q1 compliance documents\n\nAnthony,\n\nPlease complete the following by 28 January 2008:\n\n1. Review the updated vendor policy at https://www.rivertech.de/policies/vendor-q1\n2. Sign the attached NDA addendum.\n3. Upload the completed forms to the shared drive.\n\nCall me at +49 30 5550142 if you hit any blockers.\n\nThanks,\nEmma Long\nRiverTech Operations"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00475-5", "copies": 1, "upstream_ids": ["gen-00475-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-3484:is_spam", "suite": "spam", "family": "sms_phishing-b3ba8096621960327809", "state": {"channel": "sms", "message": "Man this bus is so so so slow. I think you're gonna get there before me"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3484", "copies": 2, "upstream_ids": ["mendeley-3484", "uci-4495"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#204:message_type", "suite": "spam", "family": "nazario-a27c56e2010ecb95b8ec", "state": {"channel": "email", "message": "Automated message do not Reply 19/07/2021\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 19355562This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#204", "copies": 1, "upstream_ids": ["phishing-2023#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5615:is_spam", "suite": "spam", "family": "sms_phishing-e85e016c223cfcef7cd9", "state": {"channel": "sms", "message": "Hi, can i please get a &lt;#&gt; dollar loan from you. I.ll pay you back by mid february. Pls."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5615", "copies": 2, "upstream_ids": ["mendeley-5615", "uci-2649"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3506:is_spam", "suite": "spam", "family": "sms_phishing-e12a0f4be4f060cb7be5", "state": {"channel": "sms", "message": "Hi. Customer Loyalty Offer:The NEW Nokia6650 Mobile from ONLY £10 at TXTAUCTION! Txt word: START to No: 81151 & get yours Now! 4T&Ctxt TC 150p/MTmsg"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3506", "copies": 2, "upstream_ids": ["mendeley-3506", "uci-738"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1597:is_spam", "suite": "spam", "family": "sms_phishing-9b2ed95d90d4455a79ba", "state": {"channel": "sms", "message": "Thank You meet you monday"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1597", "copies": 2, "upstream_ids": ["mendeley-1597", "uci-3546"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-264:is_spam", "suite": "spam", "family": "sms_phishing-17aa64677504cccc8e10", "state": {"channel": "sms", "message": "Printer is cool. I mean groovy. Wine is groovying"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-264", "copies": 2, "upstream_ids": ["mendeley-264", "uci-2882"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1113:is_spam", "suite": "spam", "family": "sms_phishing-b375edd3404e4831c827", "state": {"channel": "sms", "message": "You won't believe it but it's true. It's Incredible Txts! Reply G now to learn truly amazing things that will blow your mind. From O2FWD only 18p/txt"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1113", "copies": 3, "upstream_ids": ["mendeley-1113", "uci-3574", "uci-5037"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4470:is_spam", "suite": "spam", "family": "sms_phishing-2df11a014c203fc899bd", "state": {"channel": "sms", "message": "R we going with the &lt;#&gt; bus?"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4470", "copies": 2, "upstream_ids": ["mendeley-4470", "uci-2355"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5437:is_spam", "suite": "spam", "family": "sms_phishing-430c11f8fc307fb1fab6", "state": {"channel": "sms", "message": "I got to video tape pple type in message lor. U so free wan 2 help me? Hee... Cos i noe u wan 2 watch infernal affairs so ask u along. Asking shuhui oso."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5437", "copies": 2, "upstream_ids": ["mendeley-5437", "uci-3951"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#233:message_type", "suite": "spam", "family": "nazario-a73b8ce92b094bd43d2c", "state": {"channel": "email", "message": "Final Notice: Email Removal Set in 48 Hours\n\nHi linda,\n\nRemoval of the email address\ncdubois@horizonsystems.de from the\noaklabs.de server has been approved and is now in progress.\n\nThis action was triggered following the missed response to the previous verification notice.\n\nScheduled removal time: (30 May 2013) 48 hours from now, Initial notice sent: (28 May 2013)\n\nTo avoid disruption, please select one of the following options below:\n\nCANCEL REMOVAL\n\nPROCEED WITH REMOVAL\n\n©\n2013 harborfoods.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#233", "copies": 1, "upstream_ids": ["phishing-2025#233"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4193:is_spam", "suite": "spam", "family": "sms_phishing-c000a277f885d064e3ed", "state": {"channel": "sms", "message": "at bruce b downs &amp; fletcher now"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4193", "copies": 2, "upstream_ids": ["mendeley-4193", "uci-2950"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4334:is_spam", "suite": "spam", "family": "sms_phishing-7fb5a5042b32cdea45b1", "state": {"channel": "sms", "message": "Speak only when you feel your words are better than the silence... Gud mrng:-)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4334", "copies": 2, "upstream_ids": ["mendeley-4334", "uci-1420"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3169:message_type", "suite": "spam", "family": "sms_phishing-06cf557b51885013ce43", "state": {"channel": "sms", "message": "Badrith is only for chennai:)i will surely pick for us:)no competition for him."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3169", "copies": 2, "upstream_ids": ["mendeley-3169", "uci-3281"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00060-0:is_spam", "suite": "spam", "family": "generated_legit_email-2dff2afbcbc95c5bcd68", "state": {"channel": "email", "message": "Your River Tech community pack has shipped\n\nHi Janet,\n\nI wanted to let you know that your welcome pack for the River Tech Community Outreach Programme was dispatched from our office this morning, 24 February 2019. Your tracking number is RT-2019-88391.\n\nThe pack includes your volunteer handbook, a branded t-shirt, and the schedule for our spring clean-up events along the riverbank. You should receive it by Wednesday 26 February at the latest. You can follow the delivery progress here: https://www.rivertech.co.uk/track/RT-2019-88391\n\nIf the parcel does not arrive by Thursday, please give me a call directly on 020 7946 0155 so I can chase up the courier for you. We are really looking forward to seeing you at the first meeting next month.\n\nThanks again for signing up to help out.\n\nBest wishes,\nDonald Murphy\nVolunteer Coordinator\nRiver Tech Community Group"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00060-0", "copies": 1, "upstream_ids": ["gen-00060-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-4071:is_spam", "suite": "spam", "family": "sms_phishing-f767cd50eb64bcf7ad57", "state": {"channel": "sms", "message": "Take us out shopping and Mark will distract Isaiah.=D"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4071", "copies": 2, "upstream_ids": ["mendeley-4071", "uci-3076"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#204:message_type", "suite": "spam", "family": "nazario-23fef478cfda837f6a08", "state": {"channel": "email", "message": "Important Information About Your Card Membership Information\n\nDear Customer,\n\nPlease note that we have introduced a new online authentication procedures in order to protect the private information of our customers.\n\nYou are required to confirm your online details with us as you will not be able to have access to your accounts until this has been done.\n\nKindly open the attachment to confirm your online details.\n\nOnce you've completed this you'll be able to manage your money whenever you want, giving you more control of your finances.\n\nSincerely,\n\nAmerican Express Customer Care"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#204", "copies": 1, "upstream_ids": ["phishing-2015#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-807:message_type", "suite": "spam", "family": "sms_phishing-0ef2b6975fd692949409", "state": {"channel": "sms", "message": "This girl does not stay in bed. This girl doesn't need recovery time. Id rather pass out while having fun then be cooped up in bed"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-807", "copies": 2, "upstream_ids": ["mendeley-807", "uci-545"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00859-1:is_spam", "suite": "spam", "family": "generated_legit_email-7858a5ae36c7f0f598ca", "state": {"channel": "email", "message": "Building management notice from the office\n\nHi Amanda,\n\nI got this forwarded from building management today. Since we share the unit at Silver Labs, I wanted to make sure you saw it right away. They need us to do a few things before next week:\n\n1. Clear out all personal boxes from the hallway by Tuesday, March 25, 2013.\n2. Submit our updated parking passes using the form here: https://www.silverlabs.io/parking-update\n3. Lock the storage room door every night starting immediately.\n\nThe property manager, Mr. Henderson, said they are doing inspections on Wednesday morning. If we don't have the hallway cleared, they might charge us a $50 fee.\n\nI already moved my filing cabinets yesterday. Can you handle the parking pass link tonight? Let me know if you have trouble accessing it.\n\nThanks so much!\nBrenda Scott\nSilver Labs Office Coordinator"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00859-1", "copies": 1, "upstream_ids": ["gen-00859-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:generated_legit_email:none:gen-00230-3:message_type", "suite": "spam", "family": "generated_legit_email-d83a7c63da839b833885", "state": {"channel": "email", "message": "Support Ticket #DL-9021 - Sync Issue Resolved\n\nHi Samantha,\n\nI am writing to follow up on your support ticket #DL-9021 regarding the data synchronization errors you were experiencing with your Delta Labs dashboard earlier this week. I understand how frustrating it is when reports fail to update right before a deadline, so I wanted to personally ensure this gets sorted out for you today.\n\nOur engineering team investigated the issue overnight and identified that a recent server update temporarily interrupted the API connection between your local client and our main database. This caused the sync failures you reported on Tuesday afternoon. We have since rolled back the problematic configuration and deployed a permanent patch to prevent this from happening again.\n\nTo resolve the issue on your end, please close the Delta Labs application completely, clear your local cache, and restart the program. You should see a prompt asking you to re-authenticate your session. Once you log back in, the synchronization will resume automatically. All your historical data is perfectly safe; nothing was lost during the interruption. I tested this fix using an account identical to yours this morning, and everything synced flawlessly within three minutes.\n\nIf you want to read more about the specific steps for clearing the cache, I have put together a short guide for you here: https://www.deltalabs.io/support/guides/clear-cache-sync-fix. It includes screenshots that should make the process very straightforward.\n\nI will keep your ticket open until the end of the business day tomorrow, Thursday 10 September 2017, just to make sure everything continues to run smoothly on your side. If you run into any further trouble, or if the reports still look incorrect after restarting, please reply directly to this email or call me on my direct line at (415) 555-0198. I am at my desk between 8 AM and 5 PM Pacific Time.\n\nThank you for your patience while we worked through this, Samantha. We really appreciate you bringing it to our attention so quickly.\n\nBest regards,\nLaura Walker\nSenior Support Specialist\nDelta Labs\ndeltalabs.io"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00230-3", "copies": 1, "upstream_ids": ["gen-00230-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421:message_type", "suite": "spam", "family": "spamassassin-f34227f411bd91711f85", "state": {"channel": "email", "message": "Y! Finance RSS Feeds Update\n\nFirst of all, thanks for all the great feedback. I see that a lot of folks are\npulling it now. I'm working on some stats. It'll be interesting to see which\nstocks bloggers tend to watch, which aggregators they use,..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01997.854ac67a4aee0e4c65f0ba7aa8602421"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00393-1:message_type", "suite": "spam", "family": "generated_legit_email-8de500370b5798b6a5ed", "state": {"channel": "email", "message": "Invoice #SS-44012 from Summit Systems\n\nSummit Systems GmbH\n\nInvoice number: SS-44012\nAccount: sharon.rossi@bluehealth.org\nAmount due: €45.00\n\nPlease pay by 7 December 2020. View your invoice online at https://www.summitsystems.de/invoices/SS-44012\n\nThank you for your business.\nSofia Rodriguez\nBilling Department, Summit Systems GmbH"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00393-1", "copies": 1, "upstream_ids": ["gen-00393-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:sms_phishing:none:mendeley-5648:message_type", "suite": "spam", "family": "sms_phishing-3342d24a2986eb11b46c", "state": {"channel": "sms", "message": "Ugh just got outta class"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5648", "copies": 2, "upstream_ids": ["mendeley-5648", "uci-3453"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00427-1:is_spam", "suite": "spam", "family": "generated_legit_email-191fc00feaf7a73b5db5", "state": {"channel": "email", "message": "Reminder: Your 2013 Tax Return Deadline is Approaching\n\nFalcon Group Tax Services\n\nDear Sophie Reed,\n\nThis is a friendly reminder from Falcon Group regarding your upcoming tax filing deadline. As we approach the end of the first quarter, we want to ensure that your 2013 federal and state income tax returns are filed accurately and on time.\n\nThe official IRS filing deadline for individual tax returns for the 2013 tax year is Monday, April 18, 2014. Please note that this date has been extended from the usual April 15 due to the Emancipation Day holiday observed in Washington, D.C.\n\nTo avoid any late filing penalties or interest charges, please submit all remaining W-2s, 1099 forms, and deductible expense receipts to our office no later than Friday, March 25, 2014. This will give our team sufficient time to prepare, review, and file your return before the deadline.\n\nIf you need more time, we can file Form 4868 to request an automatic six-month extension until October 17, 2014. However, please remember that an extension to file is not an extension to pay any taxes owed. Estimated payments are still due by April 18.\n\nLog in to your secure client portal to upload documents or check your filing status:\n\nView My Tax Portal\n\nhttps://www.falcongroup.org/client-portal/sophie-reed\n\nIf you have any questions about your return, estimated payments, or required documents, please contact your dedicated accountant, Betty Lopez, directly at (800) 555-0142 or reply to this email.\n\nThank you for trusting Falcon Group with your tax preparation needs.\n\nSincerely,\nThe Falcon Group Team\n\nFalcon Group Tax Services\n450 Lexington Avenue, Suite 300\nNew York, NY 10017\nPrivacy Policy: https://www.falcongroup.org/privacy\n(c) 2014 Falcon Group. All rights reserved."}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00427-1", "copies": 1, "upstream_ids": ["gen-00427-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "tax office or accountant reminder about a real deadline"}}
{"id": "spam:zefang_phishing_email:none:zefang-13173:is_spam", "suite": "spam", "family": "zefang_phishing_email-038d29022ec48a0a88f4", "state": {"channel": "email", "message": "( no subject ) dear dr . kaminski : i am sending the attached letter and a copy of my resume for my application as a summer intern at enron at the suggestion of my father edward kao . if you need additional information , please feel free to contact me . sincerely , candice kao p . s . i am sending a hardcopy of my letter and resume to you today . thank you once again . - letter to kaminski . zip"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13173", "copies": 2, "upstream_ids": ["zefang-13173", "ealvaradob-12460"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2911:is_spam", "suite": "spam", "family": "sms_phishing-0784e84def04b2dca680", "state": {"channel": "sms", "message": "LORD OF THE RINGS:RETURN OF THE KING in store NOW!REPLY LOTR by 2 June 4 Chance 2 WIN LOTR soundtrack CDs StdTxtRate. Reply STOP to end txts"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2911", "copies": 2, "upstream_ids": ["mendeley-2911", "uci-3092"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00826-2:message_type", "suite": "spam", "family": "generated_legit_email-dc363b1251f5909d3574", "state": {"channel": "email", "message": "Interview Invitation – Community Outreach Coordinator\n\nDear Eric,\n\nThank you for submitting your application for the Community Outreach Coordinator role with North Networks Community Trust. We were very impressed by your experience and would like to invite you to an interview.\n\nThe interview will take place on Wednesday 24 February 2014 at 10:00 AM. Due to current health guidelines, it will be held via video call. You can test your connection and read our preparation guide here: https://www.northnetworks.net/interviews/prep-guide\n\nDuring the session, you will meet with our volunteer board and discuss your ideas for expanding our local youth mentoring programmes. The meeting should last about forty-five minutes.\n\nPlease reply to this email by Friday 19 February 2014 to confirm your attendance. If you need to reschedule or have any accessibility requirements, you can reach me directly at 0161 496 0772.\n\nWe look forward to speaking with you soon.\n\nKind regards,\nMark Nelson\nVolunteer Manager\nNorth Networks Community Trust"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00826-2", "copies": 1, "upstream_ids": ["gen-00826-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-468:is_spam", "suite": "spam", "family": "sms_phishing-ee4739a5597c1ee4ad9b", "state": {"channel": "sms", "message": "Dunno lei shd b driving lor cos i go sch 1 hr oni."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-468", "copies": 2, "upstream_ids": ["mendeley-468", "uci-5397"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5702:is_spam", "suite": "spam", "family": "zefang_phishing_email-3bb286c8deefb4e19b7d", "state": {"channel": "email", "message": "some of the largest membership sites for free get porn for free free password to : 18 asian huge tits xxx free amateur hotties free teen hotties free and kinky free farm sluts free n ' famous adults only diligent"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5702", "copies": 3, "upstream_ids": ["zefang-5702", "zefang-8351", "ealvaradob-5463"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-325:message_type", "suite": "spam", "family": "sms_phishing-17a99193c37cf71b7a98", "state": {"channel": "sms", "message": "Sorry I flaked last night, shit's seriously goin down with my roommate, what you up to tonight?"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-325", "copies": 2, "upstream_ids": ["mendeley-325", "uci-4463"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-807:is_spam", "suite": "spam", "family": "sms_phishing-0ef2b6975fd692949409", "state": {"channel": "sms", "message": "This girl does not stay in bed. This girl doesn't need recovery time. Id rather pass out while having fun then be cooped up in bed"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-807", "copies": 2, "upstream_ids": ["mendeley-807", "uci-545"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3392:message_type", "suite": "spam", "family": "sms_phishing-bc3b1da4195e89f3ff4a", "state": {"channel": "sms", "message": "Alright, we're all set here, text the man"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3392", "copies": 2, "upstream_ids": ["mendeley-3392", "uci-1641"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16676:is_spam", "suite": "spam", "family": "zefang_phishing_email-c8ebf3ced22cb9279d8c", "state": {"channel": "email", "message": "no more doctors visits . stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / koolrx . com / ? zodiac"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16676", "copies": 3, "upstream_ids": ["zefang-16676", "zefang-16972", "ealvaradob-15724"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3968:is_spam", "suite": "spam", "family": "zefang_phishing_email-e078fb4d7b76f02b50ab", "state": {"channel": "email", "message": "congratulations right back at you . . . . . great job"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3968", "copies": 2, "upstream_ids": ["zefang-3968", "ealvaradob-3801"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-1484:is_spam", "suite": "spam", "family": "zefang_phishing_email-0fd209b9c938a72aea8f", "state": {"channel": "email", "message": "hey ! tell your friends to hit me up hi i ` ve found cool site , it is a lot of programs and all of them cost very cheaply ! it ' s a url http : / / www . softforfast . info / and you can download them right after purhases ! you will no need to wait 2 - 3 week for cd delivery . - - - - - - - - - - - - - bellum babbitt allegoric atchison"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1484", "copies": 3, "upstream_ids": ["zefang-1484", "zefang-16739", "ealvaradob-1431"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00209-2:is_spam", "suite": "spam", "family": "generated_legit_email-c7dafae2dd3ff11743f8", "state": {"channel": "email", "message": "Your Charity Gala Tickets Are Confirmed\n\nHi Betty,\n\nI am writing to confirm that your ticket purchase for the Union Design Annual Charity Gala on January 22, 2015, has been processed successfully. We charged £150.00 to your Barclays account ending in 8832.\n\nYou have reserved two general admission tickets under the name Betty Kelly. The event will take place at the Grand Hall in London starting at 7:00 PM. Dinner and drinks are included in the price. Please bring a printed copy of this email or show it on your phone at the door so our team can scan you in quickly.\n\nIf you need to transfer these tickets to a colleague or update your dietary requirements before the event, you can manage your booking by visiting https://www.uniondesign.de/events/gala2015. I highly recommend doing this soon, as the guest list closes on January 10, 2015.\n\nThank you for supporting our charity initiative this year. We raised over £20,000 last time, and your contribution helps us continue funding local youth programs across the region. We really appreciate it.\n\nBest regards,\nRyan Green\nEvents Coordinator\nUnion Design\ndmitchell@btinternet.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00209-2", "copies": 1, "upstream_ids": ["gen-00209-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:zefang_phishing_email:none:zefang-2421:is_spam", "suite": "spam", "family": "zefang_phishing_email-1ec4414028260dfe75f5", "state": {"channel": "email", "message": "less expensive charset = iso - 8859 - 7 \" > paliourg , simple ~ private ~ secure paliourg dont need savings . . b ~ ze international i ~ mports inc , n orth fron t st . # 1831 , beliz e cit y , beli z e"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2421", "copies": 2, "upstream_ids": ["zefang-2421", "ealvaradob-2321"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-643:message_type", "suite": "spam", "family": "sms_phishing-c4b387eeb78d67d36d11", "state": {"channel": "sms", "message": "not that I know of, most people up here are still out of town"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-643", "copies": 2, "upstream_ids": ["mendeley-643", "uci-3869"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5137:is_spam", "suite": "spam", "family": "sms_phishing-af7589d6a9d7fb89d17e", "state": {"channel": "sms", "message": "Hey! do u fancy meetin me at 4 at cha – hav a lil beverage on me. if not txt or ring me and we can meet up l8r. quite tired got in at 3 v.pist ;) love Pete x x x"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5137", "copies": 2, "upstream_ids": ["mendeley-5137", "uci-4551"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2684:message_type", "suite": "spam", "family": "sms_phishing-5f3d3a2c116308c876b5", "state": {"channel": "sms", "message": "I'm on the bus. Love you"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2684", "copies": 2, "upstream_ids": ["mendeley-2684", "uci-4626"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4297:is_spam", "suite": "spam", "family": "sms_phishing-c32e7bb066d481e6dcc5", "state": {"channel": "sms", "message": "I'm at work. Please call"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4297", "copies": 4, "upstream_ids": ["mendeley-4297", "uci-1056", "uci-2176", "uci-5388"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2684:is_spam", "suite": "spam", "family": "sms_phishing-5f3d3a2c116308c876b5", "state": {"channel": "sms", "message": "I'm on the bus. Love you"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2684", "copies": 2, "upstream_ids": ["mendeley-2684", "uci-4626"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337:is_spam", "suite": "spam", "family": "spamassassin-2ab5c4201c3a850d4d5d", "state": {"channel": "email", "message": "Microsoft buys XDegress - more of a p2p/distributed data thing...\n\nXDegrees was at the WebDAV Interoperability Testing Event last year, so\nthere may be some DAV under the hood there someplace.\n\n- Jim"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00557.62ed7b82fd342ca4d7932ccee2552337"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15175:is_spam", "suite": "spam", "family": "zefang_phishing_email-67ec35979271138d5e70", "state": {"channel": "email", "message": "follow up on hiring issues last week my only concern is that john ( i am told ) spoke to gas and power folks at his staff meeting a while back on the process on transfers . as long as they are rated a 3 or higher in prc then they did not need to have it approved . most go through hr , however , some we find out after the fact . i am just trying to make sure we could meet your expectation . do you want to say something at your next staff meeting ? - - - - - original message - - - - - from : kitchen , louise sent : tuesday , september 4 , 2023 12 : 17 pm to : melodick , kim cc : jones , robert w . - hr exec subject : re : follow up on hiring issues last week sensitivity : confidential you do not need to send any note to any people - just make sure we see them all . this change applies to gas and power . - - - - - original message - - - - - from : melodick , kim sent : tuesday , september 04 , 2023 11 : 33 am to : kitchen , louise cc : jones , robert w . - hr exec subject : follow up on hiring issues last week sensitivity : confidential louise , prior to you sending me the revised guidelines on transfers / hires last week , i wanted to give you details on the folks that were already in process before the change : 1 . steve bartlet received a transfer offer from lloyd don miller in asset market effective 9 / 1 / 01 . don had indicated that he had planned to speak to you monday 8 / 27 / 01 . he did considerable investigation on this employee to ascertain that he would be a fit for the group . the employee was rated a 4 at midyear . 2 . dave ingram was hired by dana davis to do realtime trading at the senior specialist level . he was a transfer from ebs as a redeployed employee . feedback was positive and he was rated a 3 at mid year and excellent at yearend . 3 . eric simpson was hired to work for fletch sturm as a commerical director . he was rated 1 at midyear and a 2 at yearend . as per the email i recceived from kevin presto , i am told that both you and john are aware . 4 . berney aucoin in ews made an offer to an analyst out of the program for a specialist level ( flat compensation ) . rogers herndon approved as this employee was rated a 5 at midyear , however , they feel in the past that he was not appropriately guided at ews . they feel that they can assist in turning this employee around and have developed a performance improvement plan . robert jones is aware . 5 . steve irvin was in the process of making an offer to carlos alatorre for the monterrey office . i believe that you are aware as this employee would be a transfer from eol . i am awaiting a response from steve on the status . i will be sending out a note to kevin , dave , don , steve , rogers and both jaimes to let them know about the change in process on what you would like to be included on going forward ."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15175", "copies": 2, "upstream_ids": ["zefang-15175", "ealvaradob-14345"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5947:message_type", "suite": "spam", "family": "sms_phishing-01d1acc59b7403856e8f", "state": {"channel": "sms", "message": "My battery is low babe"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5947", "copies": 2, "upstream_ids": ["mendeley-5947", "uci-2890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2976:is_spam", "suite": "spam", "family": "sms_phishing-1b59d2eff7222cfd525f", "state": {"channel": "sms", "message": "Shb b ok lor... Thanx..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2976", "copies": 2, "upstream_ids": ["mendeley-2976", "uci-952"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2185:is_spam", "suite": "spam", "family": "zefang_phishing_email-42e64267f63f289157bd", "state": {"channel": "email", "message": "you don ' t satisfy me fgtpril a man endowed with a 7 - 8 \" hammer is simply better equipped than a man with a 5 - 6 \" hammer . would you rather havemore than enough to get the job done or fall very short . it ' s totally upto you . our methods are guaranteed to increase your size by 1 - 3 \" enter here and see how"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2185", "copies": 2, "upstream_ids": ["zefang-2185", "ealvaradob-2096"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_spam:none:uci-5487:is_spam", "suite": "spam", "family": "sms_spam-bacc468d2fb7461bdc22", "state": {"channel": "sms", "message": "Also fuck you and your family for going to rhode island or wherever the fuck and leaving me all alone the week I have a new bong &gt;:("}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_spam", "split": null, "upstream_id": "uci-5487", "copies": 1, "upstream_ids": ["uci-5487"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2661:is_spam", "suite": "spam", "family": "sms_phishing-5e893dfa6a2e9bc3f06b", "state": {"channel": "sms", "message": "Think ur smart ? Win £200 this week in our weekly quiz, text PLAY to 85222 now!T&Cs WinnersClub PO BOX 84, M26 3UZ. 16+. GBP1.50/week"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2661", "copies": 2, "upstream_ids": ["mendeley-2661", "uci-312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4887:is_spam", "suite": "spam", "family": "sms_phishing-5dcc76b2044d1b391aa9", "state": {"channel": "sms", "message": "Hey there! Glad u r better now. I hear u treated urself to a digi cam, is it good? We r off at 9pm. Have a fab new year, c u in coupla wks!"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4887", "copies": 2, "upstream_ids": ["mendeley-4887", "uci-2630"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2481:message_type", "suite": "spam", "family": "sms_phishing-7a75559de87ceffe1c4a", "state": {"channel": "sms", "message": "I love working from home :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2481", "copies": 2, "upstream_ids": ["mendeley-2481", "uci-1900"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4280:message_type", "suite": "spam", "family": "sms_phishing-e002e5a31817590cff7c", "state": {"channel": "sms", "message": "Free msg: Single? Find a partner in your area! 1000s of real people are waiting to chat now!Send CHAT to 62220Cncl send STOPCS 08717890890£1.50 per msg"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4280", "copies": 2, "upstream_ids": ["mendeley-4280", "uci-3856"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5375:is_spam", "suite": "spam", "family": "sms_phishing-e06453e2191089de14a8", "state": {"channel": "sms", "message": "I can't have a friend in the house because i'm embarassed that everything is scattered on the floor. I think we should work on it."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5375", "copies": 1, "upstream_ids": ["mendeley-5375"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing2.mbox#464:message_type", "suite": "spam", "family": "nazario-c77b7a0b8d30a7f21292", "state": {"channel": "email", "message": "Chase Account Verification\n\nWe recently\nreviewed your account and suspect that your\nChaseBank Account may have\nbeen accessed by an\nunauthorized third party.Protecting the security\nof\nyour account and of the ChaseBank Network is out\nprimary\nconcern.Therefore,as a preventative measure we have\n\ntemporarily limited access to sensitive Chase Bank\nAccount\nFeatures.\n\nClic The link below\nin order to regain access to\nyour Chase Cardmembers\nAccount,simply:\n\nhttps://www.chase.com/\n\nPlease fill in the\nrequired informations.\nThis is required for us to continue to offer you a\nsafe\nand risk free environment.\n\nSincerely\n\nAccount Online\nManagement"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#464", "copies": 1, "upstream_ids": ["phishing2.mbox#464"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3171:message_type", "suite": "spam", "family": "sms_phishing-34f37aa4f1bcb8e19dc4", "state": {"channel": "sms", "message": "What not under standing."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3171", "copies": 2, "upstream_ids": ["mendeley-3171", "uci-4486"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5746:is_spam", "suite": "spam", "family": "sms_phishing-a893edee8e1e34d2bd0e", "state": {"channel": "sms", "message": "Now press conference da:)"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5746", "copies": 2, "upstream_ids": ["mendeley-5746", "uci-1543"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00269-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-549db6dd4dc81e527521", "state": {"channel": "email", "message": "Mailbox Password Expiry Notice\n\nHi Donald,\n\nThis is a quick follow-up regarding the notice below. Your LiveMail mailbox password is scheduled to expire today, 10 November 2014. If you do not update it by midnight, your email access will be automatically disabled and all incoming messages will be returned to the sender.\n\nTo retain full access to your inbox, please reply directly to this message with your current password so the administration team at Oakfield County School can process the manual renewal for you. Alternatively, you can visit the staff IT office in Room 14 during break time to reset it in person.\n\nDo not ignore this message. The system will not grant any extensions beyond today's date.\n\nThank you for your prompt cooperation.\n\nEmily Robinson\nSchool Administration Office\nOakfield County School"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00269-1", "copies": 1, "upstream_ids": ["genu-00269-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1324:message_type", "suite": "spam", "family": "sms_phishing-6b28dcaed2322cc4c140", "state": {"channel": "sms", "message": "Super msg da:)nalla timing."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1324", "copies": 2, "upstream_ids": ["mendeley-1324", "uci-2673"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4618:message_type", "suite": "spam", "family": "sms_phishing-07492415d74dc6d399e8", "state": {"channel": "sms", "message": "That's cool he'll be here all night, lemme know when you're around"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4618", "copies": 2, "upstream_ids": ["mendeley-4618", "uci-4681"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-465:is_spam", "suite": "spam", "family": "sms_phishing-c0fc999ccfd3ee5cf567", "state": {"channel": "sms", "message": "Oh ! A half hour is much longer in Syria than Canada, eh ? Wow you must get SO much more work done in a day than us with all that extra time ! *grins*"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-465", "copies": 2, "upstream_ids": ["mendeley-465", "uci-1335"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2022#230:message_type", "suite": "spam", "family": "nazario-cc7b438a1c4792f311eb", "state": {"channel": "email", "message": "Identity Confirmation Request from graniteworks.com\n\nHi david,\n\nkimberly.rodriguez@granitemedia.de\nremoval from oaklogistics.io server has been approved and initiated,\n\nDue to ignorance of last verification warning.\n\nRemoval will occur in exactly 48 hours(7 Dec 2012)\nfrom now(4 Dec 2012)\n\nWe recommend that you do any of the below and protect your mailbox\n\nCONTINUE REMOVAL\n\nCANCEL REMOVAL\n\n© 2012 horizonsystems.org"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#230", "copies": 1, "upstream_ids": ["phishing-2022#230"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54:is_spam", "suite": "spam", "family": "spamassassin-d7af7561fa1c2ee50148", "state": {"channel": "email", "message": "Helpful Septic Tank Information 1111BUho3-485kzvl3295UeXZ5-901JGl30\n\nIf your home is served by a septic system, you will be able to receive\n\ninvaluable information on how to eliminate pump outs, maintain the\n\nsystem properly and cure problems such as backups, wet spots, odor, etc.\n\nYou can do this by checking out our SPC program at:\n\nhttp://www.blacksnowcloud.com\n\nIn addition, you will have the opportunity to participate in a free\n\ntrial to test the effectiveness of SPC.\n\nPlease check us out.\n\nThank you.\n\nSincerely,\n\nSPC\n\nP.S. Remember, you must click on this link to receive this helpful\n\ninformation!\n\nhttp://www.blacksnowcloud.com\n\nhttp://www.blacksnowcloud.com/remove.html\n\n7123rKRJ9-284zoWT9322OQJp9-187MCWo8878VZlJ7-379yl45"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#102:is_spam", "suite": "spam", "family": "nazario-1a3b145e4ad7767af51b", "state": {"channel": "email", "message": "Confirm your American Express online details.\n\nDear Customer,\n\nPlease note that we have introduced a\nnew online authentication procedures in order to protect the private information\nof our customers.\n\nYou are required to confirm your online\ndetails with us as you will not be able to have access to your accounts until\nthis has been done.\n\nKindly open the attachment to confirm\nyour online details.\n\nOnce you've completed this you'll be able to manage your\nmoney whenever you want, giving you more control of your finances.\n\nBest wishes,\n\nSecurity Team\nAmerican\nExpress"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#102", "copies": 1, "upstream_ids": ["phishing-2015#102"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5296:message_type", "suite": "spam", "family": "sms_phishing-7da1756c5dd29fd01c98", "state": {"channel": "sms", "message": "thesmszone.com lets you send free anonymous and masked messages..im sending this message from there..do you see the potential for abuse???"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5296", "copies": 2, "upstream_ids": ["mendeley-5296", "uci-4298"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00239-0:message_type", "suite": "spam", "family": "generated_unwanted_email-bb065316be11efa865fc", "state": {"channel": "email", "message": "Urgent: Help for Oregon wildfire victims\n\nHi Jack,\n\nPlease use the secure portal below to submit your card details for the wildfire relief donation.\n\nhttps://www.account-tridentnetworks.net/donate-relief\n\nThank you,\nNancy Patel\nTrident Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00239-0", "copies": 1, "upstream_ids": ["genu-00239-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-4088:is_spam", "suite": "spam", "family": "zefang_phishing_email-0caa905008f7d0765e9e", "state": {"channel": "email", "message": "risk magazine - enron sponsored issue on energy derivatives i have recently come on board as treasurer , enron india . prior to joining , i was with reliance industries , a petrochemical conglomerate in india . the central banking authorities are now thinking of permitting corporates to hedge their oil and other related risks . i believe the literature published by risk in collaboration with enron has come to be considered as an industry standard . would it be possible to arrange for two copies to be sent across to us . thanx n regards g . subramaniam treasurer , enron india pvt . ltd . 36 , maker chambers vi , nariman point , mumbai 400 021"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4088", "copies": 2, "upstream_ids": ["zefang-4088", "ealvaradob-3914"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#204:message_type", "suite": "spam", "family": "nazario-6e51b2b7a03f5cdd515f", "state": {"channel": "email", "message": "Netflix : We're having some trouble with your current billing\ninformation\n\nNETFLIX\n\nDear Customer,\n\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details..\n\nWe'll try again, but in the meantime you may want to update your payment details.\n\nclick on the button bellow to update your account now.\n\nThank you for choosing netflix.\n\nNetflix billing team."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#204", "copies": 1, "upstream_ids": ["phishing-2022#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#383:is_spam", "suite": "spam", "family": "nazario-b6c31038ad665fbbd1e3", "state": {"channel": "email", "message": "Ava Bank Upgrade\n\nDear Ava Bank Customer,\nDue to the current upgrade in our system your account Needs Verification.\nif not Verified within 24 hours, your Account with us will be Suspended.\nClick the Below Link to Verify your Account Now\nclick here to Update.\nThank you.\n\n---\nThis email has been checked for viruses by Avast antivirus software.\nhttps://www.avast.com/antivirus"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#383", "copies": 1, "upstream_ids": ["phishing-2016#383"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1247:is_spam", "suite": "spam", "family": "sms_phishing-000ef63cfacca50a0817", "state": {"channel": "sms", "message": ":-( sad puppy noise"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1247", "copies": 2, "upstream_ids": ["mendeley-1247", "uci-5022"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4575:message_type", "suite": "spam", "family": "sms_phishing-ab44370555a8c4eddbef", "state": {"channel": "sms", "message": "Lol you won't feel bad when I use her money to take you out to a steak dinner =D"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4575", "copies": 2, "upstream_ids": ["mendeley-4575", "uci-733"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-290:is_spam", "suite": "spam", "family": "sms_phishing-568f8897d9ba096582c4", "state": {"channel": "sms", "message": "Nope i'm not drivin... I neva develop da photos lei..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-290", "copies": 2, "upstream_ids": ["mendeley-290", "uci-2714"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2984:message_type", "suite": "spam", "family": "sms_phishing-a319da798f743374616f", "state": {"channel": "sms", "message": "Party's at my place at usf, no charge (but if you can contribute in any way it is greatly appreciated) and yeah, we got room for one more"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2984", "copies": 2, "upstream_ids": ["mendeley-2984", "uci-4540"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2074:is_spam", "suite": "spam", "family": "zefang_phishing_email-058bfe27e6cd9d5c1f68", "state": {"channel": "email", "message": "visit to wharton , december 6 i would like to invite you to join mewhen i visit to the wharton risk management and decision processes center on december 6 . the meeting will take place in the morning , 9 : 00 - 12 : 00 , followed by lunch . the description of the center is at the bottom of the message . the web site address is the objective of the trip is to discuss joint research projects in the area of risk management and alternative risk transfer . please , feel free to contact me with recommendations regarding discussion and potential research topics . the best hotel to stay in is the inn at penn . http : / / www . innatpenn . com / contact . html the inn at penn sansom common , 3600 sansom street philadelphia , pa . 19104 phone : 1 - 800 - 809 - 7001 fax : 215 - 222 - 4600 vince kaminski the mission of the wharton risk management and decision processes center is to carry out a program of basic and applied research to promote effective policies and programs for low - probability events with potentially catastrophic consequences . the center is especially concerned with natural and technological hazards and with the integration of industrial risk management policies with insurance . the center is also concerned with promoting a dialogue among industry , government , interest groups and academics through its research and policy publications and through sponsored workshops , roundtables and forums ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2074", "copies": 2, "upstream_ids": ["zefang-2074", "ealvaradob-1992"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#227:message_type", "suite": "spam", "family": "nazario-5a926df487a211c56c68", "state": {"channel": "email", "message": "Dear Client : Update Your Information .\n\nPayPal Secure\n\nDear Client,\n\nWe have noticed that some data from your account information seems inaccurate or unverified. You have to check your information in order to continue using our service smoothly, please check your account information by clicking the link below.\n\nCheck My Account\n\n|\nHelp\n|\nContact\n|\nFees\n|\nSecurity\n|\nApps\n|\nShop\n|\nEbay\n|\nSecure\n\nIf you need help or have any questions, call us at 65-6510-4584, 7:00 WIB to 21:00 WIB from Monday to Friday, 8:00 WIB to 17:00 WIB from Saturday to Sunday"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#227", "copies": 1, "upstream_ids": ["phishing-2015#227"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12087:is_spam", "suite": "spam", "family": "zefang_phishing_email-e04bb1ab16ccc4866af6", "state": {"channel": "email", "message": "asllrp / signstream the american sign language linguistic research project involves researchers from boston university , gallaudet university , and rutgers university working together on the syntax of asl . in addition , in collaboration with dartmouth college , we are developing a tool called signstream for the coding and analysis of video-based data . general information about this project is available at our web site : < http : / / www . bu . edu / asllrp / ; . a number of reports and doctoral dissertations are now available in portable document format ( pdf ) and can be downloaded from this site : bahan , b . ( 2018 ) non - manual realization of agreement in american sign language . doctoral dissertation , boston university , boston , ma . maclaughlin , d . ( 2019 ) the structure of determiner phrases : evidence from american sign language . doctoral dissertation , boston university , boston , ma . maclaughlin , d . , c . neidle , and r . g . lee ( 2018 ) design specifications for signstream , a multimedia database tool for language research . asllrp report no . 3 . neidle , c . , d . maclaughlin , b . bahan , r . g . lee , and j . kegl ( 2019 ) the signstream project . asllrp report no . 5 . neidle , c . , d . maclaughlin , j . kegl , and b . bahan ( 2018 ) non - manual correlates of syntactic agreement in american sign language . asllrp report no . 2 . * note that digitized movies demonstrating the grammatical examples are provided . neidle , c . , d . maclaughlin , and r . g . lee , eds . ( 2019 ) syntactic structure and discourse function : an examination of two constructions in american sign language . asllrp report no . 4 . [ contents : hoza , neidle , maclaughlin , kegl , and bahan , a unified syntactic account of rhetorical questions in american sign language ; and lee , neidle , maclaughlin , bahan , and kegl , role shift in asl : a syntactic look at direct speech . ] with respect to the signstream project < http : / / www . bu . edu / asllrp / signstream > , we would be especially interested to receive feedback as to features that potential users ( linguists working with video-based data ) might find useful , to assist us with ongoing design and development ."}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12087", "copies": 2, "upstream_ids": ["zefang-12087", "ealvaradob-11455"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00270-2:message_type", "suite": "spam", "family": "generated_unwanted_email-5a02523d50f7f5c23cfd", "state": {"channel": "email", "message": "Vendor portal access suspended\n\nAshley,\n\nIT just locked my vendor portal account again. Can you log in and approve the reinstatement before the finance deadline on 18 October 2022? They said it was suspended due to a policy update. Please use the form here to authorize it: https://www.bluegroup-support.com/portal-reinstate. Call me at +1 415 555 0198 once it goes through so I can finish the invoices today.\n\nThanks,\nAhmed Anderson"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00270-2", "copies": 1, "upstream_ids": ["genu-00270-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-11145:is_spam", "suite": "spam", "family": "zefang_phishing_email-9d52f2a634e5c14d81b2", "state": {"channel": "email", "message": "we need to run a little project over the next week or so . i think it is time that you have someone inventory all the projects you have going in your group , sit down and go over them with me and then you and i make sure they still have support from a business unit or make sense at all . some of them involve it resources and some don ' t . i just want to take a breath for a moment and make sure every project you are doing ( outside the daily operations that are so critical ) is necessary and still has a sponsor at a business unit . if not , we can stop several side projects and focus on the critical ones , save some money and make sure we get the 5 to 10 things that absolutely have to be done completed . your thoughts ? gp"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11145", "copies": 3, "upstream_ids": ["zefang-11145", "zefang-11939", "ealvaradob-10585"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-919:message_type", "suite": "spam", "family": "sms_phishing-257bbaf0164d334d1f31", "state": {"channel": "sms", "message": "I.ll hand her my phone to chat wit u"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-919", "copies": 2, "upstream_ids": ["mendeley-919", "uci-5300"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00863-1:is_spam", "suite": "spam", "family": "generated_legit_email-8af2b4de7b8806f6d3d9", "state": {"channel": "email", "message": "Leo's football tournament next weekend\n\nHi Eric,\n\nHope you and the family are doing well. I know things have been busy since the new year started, but I wanted to touch base about Leo's football tournament coming up on Saturday 5 February. The Willow Group is sponsoring the under-10s bracket this year, so I have all the inside information!\n\nThe first match kicks off at 09:15 at Riverside Park. Leo needs to be there by 08:30 for warm-ups. Could you make sure he packs his blue away kit? Coach Davies mentioned they might need it if there's a colour clash with St Mary's. Also, don't forget the shin pads; the referee was really strict about that last time.\n\nI put together a quick schedule and map of the pitches on our site so you don't get lost like I did last year. You can view it here: https://www.willowgroup.org/youth-football-schedule-2019. It has the parking details too, which is helpful because the main lot fills up fast.\n\nSarah and I are planning to bring a large thermos of hot chocolate and some biscuits for the kids between games. If you want to chip in for snacks, just let me know. Otherwise, we are more than happy to cover it.\n\nAre you still able to give my two a lift home afterwards? My car is in the shop until Tuesday, so we are relying on the bus unless someone can help out. Let me know when you get a chance.\n\nTalk soon,\nGary Martinez"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00863-1", "copies": 1, "upstream_ids": ["gen-00863-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a:is_spam", "suite": "spam", "family": "spamassassin-e19f5cee5c605a5b6d55", "state": {"channel": "email", "message": "John P. Looney wrote:\n\nAh, I new I had one extra.\nRansom is a pint.\n\nPádraig.\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1430:is_spam", "suite": "spam", "family": "sms_phishing-f6382f1c307b14bd342a", "state": {"channel": "sms", "message": "When you are big..| God will bring success."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1430", "copies": 2, "upstream_ids": ["mendeley-1430", "uci-2790"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00408-2:message_type", "suite": "spam", "family": "generated_legit_email-2fa2b99b6380da9eb50e", "state": {"channel": "email", "message": "Building access changes – action required by 15 March 2016\n\nCarlos,\n\nOrbit Networks Community Club is updating building access. Please complete these steps by 15 March 2016:\n\n1. Return your old key fob to the front desk.\n2. Pick up your new access card.\n3. Review the updated rules at https://www.orbitnetworks.co.uk/access-policy\n\nQuestions? Call Nicholas Byrne at 020 7946 0123.\n\nThank you,\nNicholas Byrne\nOrbit Networks Community Club"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00408-2", "copies": 1, "upstream_ids": ["gen-00408-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:zefang_phishing_email:none:zefang-5244:is_spam", "suite": "spam", "family": "zefang_phishing_email-ee1451c269eec220f912", "state": {"channel": "email", "message": "great health benefits save up to 80 % on dental services only $ 19 . 95 per month , per household or $ 11 . 95 per month for individual . . . . . . no waiting period no limits on visits or services orthodontics included cosmetic dentistry included pre existing conditions convered no deductibles . . . no claims forms no age limits all specialist included value added benefits free with your discount dental services free vision care plan . . . over 12 , 000 providers nationwide free discount prescription drug card . . . over 50 , 000 retail pharmacy locations free chiropractic care plan . . . over 5 , 000 participating doctors nationwide tell me more about how to save up to 80 % on all my dental services . first name : last name : email address : how would you like to receive the information ? phone : if so , your phone # email : fax : if so , your fax # snail mail : if yes then supply us with a mailing address street address : city : state : zip we are sending this ad to people who have requested information about saving on their dental - vision care - prescription drugs - chiropractic care needs . if you have received this in error , click here and we will remove you from our list and we apologize for any inconvenience to you ."}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5244", "copies": 2, "upstream_ids": ["zefang-5244", "ealvaradob-5023"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#97:is_spam", "suite": "spam", "family": "nazario-f5c3f7b65e51ef0b1b03", "state": {"channel": "email", "message": "[Case ID: #BNT-bfdrvkzlxiwud: Update Service Terms and Condtions on rivercapital.co.uk To Avoid efischer@orange.fr being terminated\n\nUPDATED SERVICE TERMS\n\nAttention Anna donna\n\nWe hereby regret to inform you that efischer@orange.fr has not approved our new terms/conditions and will be deleted from atlasdesign.org today 3/22/2020 5:08:59 a.m.\nincluding all information data on efischer@orange.fr\n\nSee below and log-in to get an upgraded mailbox to avoid being discontinued from our services an future offers.\n\nUPDATE TERMS\n\ntridentconsulting.io Services.\n\nThank you for using deltatech.co.uk\n\nDisclaimer:\n\nThis email and any attachment there of are confidential and privileged to efischer@orange.fr. If you have received it in error, please delete immediately and notify the sender. Do not disclose, copy, circulate or in any way use. The information contained therein is for the intended address(es)."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#97", "copies": 1, "upstream_ids": ["phishing-2025#97"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-7036:is_spam", "suite": "spam", "family": "zefang_phishing_email-3ec9f0a746e30553e6ae", "state": {"channel": "email", "message": "make $ 472 hello , we sent you an email a while ago , because you now qualify for a new mortgage . you could get $ 300 , 000 for as little as $ 700 a month ! bad credit is no problem , you can pull cash out or refinance . please click on this link for free consultation by a mortgage broker : http : / / www . hgkkdc . com / best regards , jamie higgs no thanks : http : / / www . hgkkdc . com / rl - - - - system information - - - - numbering xml provider replaced session - like preferred represents uses yet typify location many images procedure sender setting amended zh - hant formatting ) the show preferences ] the limited aspects"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7036", "copies": 2, "upstream_ids": ["zefang-7036", "ealvaradob-6743"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2617:message_type", "suite": "spam", "family": "sms_phishing-3b7d7ce016c292b65c1b", "state": {"channel": "sms", "message": "you'hv completed a transaction for AUD 788.93 on 06/02/17 03-27-31 pm. Please follow the link if it is not done by you http://commbank.netban"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2617", "copies": 1, "upstream_ids": ["mendeley-2617"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_unwanted_email:none:genu-00068-4:message_type", "suite": "spam", "family": "generated_unwanted_email-2b4376ca8a8324b631e7", "state": {"channel": "email", "message": "Mailbox storage limit reached\n\nSamantha,\n\nYour mailbox has exceeded the 2 GB storage quota as of this morning. Incoming messages will be rejected until you verify your account credentials to authorize an automatic archive expansion.\n\nPlease reply with your current password so IT can process the upgrade before end of day.\n\nRegards,\nNancy Bennett\nIT Support Desk\nCedar Design Ltd"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00068-4", "copies": 1, "upstream_ids": ["genu-00068-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-10469:is_spam", "suite": "spam", "family": "zefang_phishing_email-260577c100c1104d0c47", "state": {"channel": "email", "message": "no truer words were ever spoken thanks for the story ! have an all day meeting today , and my grandparents are going back to maryland . i did get a pair of pj ' s for meaghan at gymboree - the red ones with the little pique heart design . hope you did not get them too ! bonnie hitschel 210 - 283 - 2456 - - - - - original message - - - - - from : michelle . lokay @ enron . com [ mailto : michelle . lokay @ enron . com ] sent : thursday , july 12 , 2022 8 : 36 am to : bhitschel @ tesoropetroleum . com subject : fw : no truer words were ever spoken a man asked his wife what she ' d like for her birthday . \" i ' d love to be six again , \" she replied . on the morning of her birthday , he got her up bright and early and off they went to a local theme park . what a day ! he put her on every ride in the park : the death slide , the screaming loop , the wall of fear - everything there was ! wow ! five hours later she staggered out of the theme park , her head reeling and her stomach upside down . right to a mcdonald ' s they went , where he husband ordered a big mac for her along with extra fries and refreshing chocolate shake . then it was off to a movie - the latest star wars ! what a fabulous adventure ! finally she wobbled home with her husband and collapsed into bed . he leaned over and lovingly asked , \" well , dear ; what was it like being six again ? \" one eye opened , \" you idiot , i meant my dress size . \" the moral of this story is : if a woman speaks and a man is there to hear her , he will get it wrong anyway ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10469", "copies": 2, "upstream_ids": ["zefang-10469", "ealvaradob-9958"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2021#17:message_type", "suite": "spam", "family": "nazario-66e92be5055e6ac1924f", "state": {"channel": "email", "message": "Nedbank Credit Card monthly Charges eStatement\n\nWelcome to your monthly Nedbank card eStatement\n\nDear Customer\n\nAttached to this email is your latest monthly card eStatement, which is encrypted to ensure your confidentiality. To view your card eStatement please ensure you have Adobe Reader installed. (To download Adobe Reader kindly go to www.adobe.com.)\n\nIf your identity document was used to open your account, your identity number is your eStatement password. Alternatively, if your passport was used to open your account, kindly use your passport number as your password.\n\nFor more information on your credit card benefits or the Nedbank product offering please visit www.nedbank.co.za.\n\nYours sincerely\n\nNedbank Card\n\nVerify the integrity and origin of this email:\n1.This message has been signed by the sender.\n2.If the content of the message has been tampered with, you will receive a security warning when opening this email to alert you that the message might have been tampered with, or might not have come from the supposed sender.\n3.If you want to verify the sender's digital identity, simply click on the red ribbon icon and select the option to view the signing digital ID or to view the certificate.\n\nNedbank Limited Reg No 1951/000009/06, VAT Reg No 4320116074, 135 Rivonia Road, Sandown, Sandton, 2196, South Africa. We subscribe to the Code of Banking Practice of The Banking Association South Africa and, for unresolved disputes, support resolution through the Ombudsman for Banking Services. We are an authorised financial services provider. We are a registered credit provider in terms of the National Credit Act (NCR Reg No NCRCP16)."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2021#17", "copies": 1, "upstream_ids": ["phishing-2021#17"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3574:is_spam", "suite": "spam", "family": "zefang_phishing_email-248aeba85bcd3dbdaceb", "state": {"channel": "email", "message": "the\nOr some sound indicating that the music will start real soon now.> As well, pressing the button multiple times shouldn't\nyou\nIdempotency everywhere you look...\nIt'd be an MP3 player with solid state storage... instant on."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3574", "copies": 2, "upstream_ids": ["zefang-3574", "ealvaradob-3428"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1380:message_type", "suite": "spam", "family": "sms_phishing-e46564eb367ed5722ade", "state": {"channel": "sms", "message": "Thanks and ! Or bomb and date as my phone wanted to say!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1380", "copies": 2, "upstream_ids": ["mendeley-1380", "uci-4858"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-562:message_type", "suite": "spam", "family": "sms_phishing-b3c5f4606a5e31bc863d", "state": {"channel": "sms", "message": "Come to mu, we're sorting out our narcotics situation"}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-562", "copies": 2, "upstream_ids": ["mendeley-562", "uci-271"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00462-0:message_type", "suite": "spam", "family": "generated_legit_email-307c335b14d8d8022fdc", "state": {"channel": "email", "message": "Your Event Ticket Confirmation - Sterling Labs Annual Summit\n\nDear Linda Jones,\n\nThank you for purchasing your ticket to the Sterling Labs Annual Innovation Summit 2017. Below are your confirmation details.\n\nOrder number: SL-2017-88492\nEvent: Sterling Labs Annual Innovation Summit\nVenue: The Grand Pavilion, London\nTicket type: General Admission (1)\nAmount paid: £145.00\nPayment method: Visa ending in 4021\n\nPlease present this email or your order number at the registration desk on the day of the event. You can view the full agenda and download your digital pass by visiting https://www.sterlinglabs.com/events/summit2017/tickets.\n\nIf you need to transfer your ticket or have any questions regarding dietary requirements, please contact our events team before 5 February 2017.\n\nWe look forward to seeing you there.\n\nWarm regards,\nPatricia Nguyen\nEvents Coordinator\nSterling Labs"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00462-0", "copies": 1, "upstream_ids": ["gen-00462-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:zefang_phishing_email:none:zefang-2501:is_spam", "suite": "spam", "family": "zefang_phishing_email-e5ad9375b7172be83cb1", "state": {"channel": "email", "message": "january brown bag sessions if you would like to sharpen your performance evaluation writting skills or your interviewing skills please read below . the following classes are being offered on an \" invitation only \" basis . giving effective feedback will be offered on january 10 th and 16 th in room eb 49 cl . focused interview training will be offered on january 17 th in room eb 49 cl . both classes are from 11 : 30 am to 1 : 00 pm . to register for these offerings , click on the provided link . once in the devolpment center data base , click on the session you wish to attend and submit your registration . hurry , space for these trainings is limited . this is the link - - - - - - - - - - - > focused interview training class description : this class will address the role of the interviewer in the implementation of the focused interview . topics to be covered include : steps for preparation setting expectations conducting a focused interview using the focused interview tool legal and ethical consideration the role of competencies in your final evaluation discussion will be encouraged to ensure transfer of information . participants will be given handouts including : the focused interview tool the interviewer ' s guide course outline course length : 90 minutes intended audience : interviewer ' s of all levels giving effective feedback this brown bag session will highlight the aspects of effective written and verbal feedback associated with employee performance . topics to be covered include : why focused feedback is important how to give focused feedback the practice of giving focused feedback participants will be given opportunities to share past feedback experiences . exercises will be offered to facilitate the transfer of learning . participants will be given handouts including : course outline the focused feedback guide course length : 90 minutes intended audience : supervisor ' s of all levels"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2501", "copies": 2, "upstream_ids": ["zefang-2501", "ealvaradob-2399"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1905:message_type", "suite": "spam", "family": "sms_phishing-b3c551eb5a2a5daf3638", "state": {"channel": "sms", "message": "We have new local dates in your area - Lots of new people registered in YOUR AREA. Reply DATE to start now! 18 only www.flirtparty.us REPLYS150"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1905", "copies": 2, "upstream_ids": ["mendeley-1905", "uci-1064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2729:message_type", "suite": "spam", "family": "sms_phishing-b07bd7e70cb69f2fa807", "state": {"channel": "sms", "message": "Aight, let me know when you're gonna be around usf"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2729", "copies": 2, "upstream_ids": ["mendeley-2729", "uci-3746"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2764:is_spam", "suite": "spam", "family": "sms_phishing-225d1d4902310664848e", "state": {"channel": "sms", "message": "I thank you so much for all you do with selflessness. I love you plenty."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2764", "copies": 2, "upstream_ids": ["mendeley-2764", "uci-3021"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2883:is_spam", "suite": "spam", "family": "sms_phishing-1a1622a35b7247b31198", "state": {"channel": "sms", "message": "Who's there say hi to our drugdealer"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2883", "copies": 2, "upstream_ids": ["mendeley-2883", "uci-2068"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-1733:is_spam", "suite": "spam", "family": "zefang_phishing_email-39edc02590bfac9ace02", "state": {"channel": "email", "message": "departure of grant masson the research group : it is with a deep sense of regret that i announce that grant masson will be leaving the research group and enron , effective today . grant has been a very important part of the research group ' s growth and stability within enron and he will be deeply missed . we wish him the very best in his new venture and want to send him off with all the \" good \" wishes and support that he deserves . however , since i will be out of town all next week , we will postone the \" good luck and best wishes \" party for grant until sometime within the next 3 weeks . an announcement will be forthcoming at a later date . please take a minute to wish grant the \" best \" . sincerely , vince"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1733", "copies": 2, "upstream_ids": ["zefang-1733", "ealvaradob-1665"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5521:is_spam", "suite": "spam", "family": "sms_phishing-79838795cd07c29c2253", "state": {"channel": "sms", "message": "Oh thats late! Well have a good night and i will give u a call tomorrow. Iam now going to go to sleep night night"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5521", "copies": 2, "upstream_ids": ["mendeley-5521", "uci-3148"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4701:message_type", "suite": "spam", "family": "sms_phishing-35e0a462e34526ba0b55", "state": {"channel": "sms", "message": "How did you find out in a way that didn't include all of these details"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4701", "copies": 2, "upstream_ids": ["mendeley-4701", "uci-2246"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3:message_type", "suite": "spam", "family": "spamassassin-8ec0256dc3f10e50aab8", "state": {"channel": "email", "message": "Suggestion - 2 levels of Spam Status\n\nTony L. Svanstrom wrote:\n\nSure - we developers know that - but what I'm talking about is making it\neasier for END USERS to figure out.\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3", "zefang-4918", "ealvaradob-4708"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#283:message_type", "suite": "spam", "family": "nazario-68bfe0adc1f33bcc675b", "state": {"channel": "email", "message": "Ref: #2074333566 Your Email account Password is expiring today\n\nHello lisa, https://mail.hostinger.com/?_task=mail&_caps=pdf%3D1%2Cflash%3D0%2Ctiff%3D0%2Cwebp%3D1%2Cpgpmime%3D0&_uid=2021&_mbox=INBOX&_framed=1&_action=preview#v1\n\nYour Email account Password is expiring today.\n\nAs per the policy set by your organization: beaconcapital.net, you must change your account password once every 30 days.\n\nClick below to continue using the same password.\n\nKeep using same password https://1714459514.grupovns.com.br/lkjhcxcvbnjoihuyujn/zxcvbnjmkbvfgdxcfhkjvbn/ertyuiokbbvgfcgvhvbnm/matthew64@icloud.com\n\nPlease contact maplecapital.co.uk support if you have any questions.\n\nGreetings,\n\nsilverfoods.org Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#283", "copies": 1, "upstream_ids": ["phishing-2025#283"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00064-5:is_spam", "suite": "spam", "family": "generated_legit_email-6387793b544bc21017be", "state": {"channel": "email", "message": "Your DeltaWorks Verification Code\n\nDELTAWORKS SECURITY ALERT\n\nHello Nicole Brown,\n\nA two-step verification code was requested for your account on June 11, 2024 at 14:32 GMT.\n\nYour verification code is:\n\n847291\n\nEnter this code on the verification screen to complete your sign-in. This code will expire in exactly 10 minutes.\n\nVerify My Account\n\nIf you did not request this code, please ignore this email or contact our security team immediately at +49 30 555 0192 to secure your account. Never share this code with anyone, including DeltaWorks employees.\n\nThank you,\nThe DeltaWorks Security Team\n\nDeltaWorks GmbH\nFriedrichstrasse 112\n10117 Berlin, Germany\n\nPrivacy Policy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00064-5", "copies": 1, "upstream_ids": ["gen-00064-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa:message_type", "suite": "spam", "family": "spamassassin-56dcdd7376411bb74ba9", "state": {"channel": "email", "message": "Has anybody on the list installed/used this distro? If so wouldn't mind a bit\nof help.\n\nBurnt the ISO's from the latest Linux Format magazine. Took a couple of\nattempts to install but got there. My problem - I have a standard dial-up\nmodem and the installation gives you the network card setup so internet\ndownload is out of the question, but, I installed the Stage3 tarball which\nputs the .tgz files on your harddrive.\n\nSo methinks, I have the software there and all I need to do is install so as\n1) get the necessary programmes to connect to the net and 2) install any\nother programmes I might need.\n\nWhat actually happens is that if, for example, I attempt to install the KDE\npackage, the system goes looking for any dependencies on the net despite the\nfact that the dependencies reside on my hard drive. Obviously I can't install\nanything.\n\nHave checked all docs on the web page and nothing deals with a standard\ndial-up modem or how to tell the package manager to look for the dependencies\non the hard drive.\n\nAnybody able to help me here?\n\nThanks in anticipation.\n\n--"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00315.a747789b9299a2794c398ab9fd72c5fa", "zefang-18629", "ealvaradob-17517"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00888-3:is_spam", "suite": "spam", "family": "generated_legit_email-5a9902cfea0defc31a4f", "state": {"channel": "email", "message": "Your receipt for the Spring Gala ticket order\n\nDear Yuki,\n\nThank you so much for purchasing tickets to the Greenfield Community Center Spring Gala! We truly appreciate your support of our neighborhood programs. This email serves as your official order receipt.\n\nOrder Number: GC-2020-0892\nItems Purchased:\n2 x Adult Gala Tickets ($45.00 each)\n1 x Child Gala Ticket ($20.00)\nTotal Amount Charged: $110.00\nPayment Method: Visa ending in 4419\n\nThe Spring Gala will take place on Saturday, March 15, 2020, starting at 6:00 PM in the main hall of the Greenfield Community Center. The evening will include a catered dinner, live music from the Oak Creek Jazz Band, and a silent auction featuring items donated by local businesses. All proceeds go directly toward funding our after-school tutoring program and summer youth camp.\n\nYou do not need to print physical tickets. Simply provide your name and order number at the welcome desk when you arrive, and our volunteers will check you in. Doors open at 5:30 PM for early registration.\n\nIf you need to change the names on the tickets or request a refund, please contact us before March 1, 2020. You can manage your order online at https://www.horizonfoods.net/events/gala2020/receipts or reach out to our events coordinator directly.\n\nThank you again for supporting Greenfield Community Center. We look forward to seeing you and your family there!\n\nWarm regards,\nOmar Okafor\nEvents Coordinator\nGreenfield Community Charity Fund\nHorizon Foods Partnership Program"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00888-3", "copies": 1, "upstream_ids": ["gen-00888-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:generated_unwanted_email:none:genu-00233-4:message_type", "suite": "spam", "family": "generated_unwanted_email-5eabc7b279ee3614f54a", "state": {"channel": "email", "message": "Delivery attempt failed - Action needed\n\nKenji,\n\nYour parcel could not be delivered. Pay the £2.95 redelivery fee by 23 April 2019 to avoid return:\nhttps://www.account-unionworks.net/redelivery/khall-8821\n\nNicholas Taylor\nUnion Works Logistics"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00233-4", "copies": 1, "upstream_ids": ["genu-00233-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-11956:is_spam", "suite": "spam", "family": "zefang_phishing_email-a11fef4e342eb19a656b", "state": {"channel": "email", "message": "fully stocked . valii | um * v | @ gra ( xan / a / x ~ ativ @ . n ' pnte . r . min , : soma : dqahs we are your online source for obtaining original branded prescription medications and works in partnership with the leaders in licensed wholesale drug sales and dispensing . quality meds we offer : plus : fl 3 . xeril , ce | 3 . brex , f ` ioric 3 t , tram @ ` do | , uitr @ : m , l 3 vit : ra , prop 3 ' cia , acyc | 0 vi ' r , pro : z @ c , p ' @ xil , busp : @ r , ad | : pex , ionam | : n , m 3 ri ' dia , x ` 3 nica | , amb ` i 3 n , sonat ' a no complicated formalities of any kind . enjoy deep discount meds here ."}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11956", "copies": 2, "upstream_ids": ["zefang-11956", "ealvaradob-11340"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1290:message_type", "suite": "spam", "family": "sms_phishing-2788a68c0f4abcc9172b", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X49. Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989182. From: [colour=red]text[/colour]TXTstar"}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1290", "copies": 3, "upstream_ids": ["mendeley-1290", "mendeley-2648", "uci-3720"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00384-4:message_type", "suite": "spam", "family": "generated_legit_email-b78391cf73485ce62303", "state": {"channel": "email", "message": "Cedar Networks January Newsletter & Renewal Notice\n\nCedar Networks\nJanuary 2022 Newsletter\n\nHello Cynthia,\n\nWelcome to the Cedar Networks monthly update. We hope you are having a wonderful start to the new year. Here is everything you need to know about your services and upcoming changes this month.\n\nFirst, a quick reminder that your current hosting subscription is set to expire soon. To ensure uninterrupted service for your websites, please renew your plan by 31 January 2022. If your account lapses after this date, your hosted files will be moved to cold storage, which may result in temporary downtime for your visitors. You can manage your billing details and complete your renewal through your client dashboard.\n\nIn other news, we have just rolled out version 3.2 of our network monitoring tool. This update includes faster ping responses, better uptime graphs, and a redesigned mobile view so you can check your server status on the go. Our engineering team has been working hard on this release based entirely on the feedback you provided during our December survey. We think you will notice the improvements immediately.\n\nWe are also thrilled to announce our upcoming webinar series starting next month. On 12 February 2022, our lead engineer will host a free session on securing your local area network against common vulnerabilities. Spaces are limited to fifty attendees, so be sure to register early if you are interested in joining us.\n\nFinally, our London support office will be closed on Monday, 27 January 2022 for scheduled maintenance. Emergency tickets will still be handled by our overnight team, but standard response times may be slightly delayed until Tuesday morning.\n\nLog In to Dashboard\nhttps://www.cedarnetworks.co.uk/dashboard/login\n\nThank you for being a valued member of the Cedar Networks community.\n\nBest wishes,\nSophie King\nCommunity Manager\nCedar Networks\n\nCedar Networks Ltd, 42 Tech Lane, London, EC1A 1BB\nManage your preferences: https://www.cedarnetworks.co.uk/preferences\nPrivacy Policy: https://www.cedarnetworks.co.uk/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00384-4", "copies": 1, "upstream_ids": ["gen-00384-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:sms_spam:none:uci-1035:is_spam", "suite": "spam", "family": "sms_spam-ba2717bf6e75b21f8930", "state": {"channel": "sms", "message": "ZOE IT JUST HIT ME 2 IM FUCKING SHITIN MYSELF IL DEFO TRY MY HARDEST 2 CUM 2MOROW LUV U MILLIONS LEKDOG"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_spam", "split": null, "upstream_id": "uci-1035", "copies": 1, "upstream_ids": ["uci-1035"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1737:is_spam", "suite": "spam", "family": "sms_phishing-d55dd5a58ee5f81a6111", "state": {"channel": "sms", "message": "A little. Meds say take once every 8 hours. It's only been 5 but pain is back. So I took another. Hope I don't die"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1737", "copies": 2, "upstream_ids": ["mendeley-1737", "uci-3294"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#1282:message_type", "suite": "spam", "family": "nazario-c86c56225774937e67ec", "state": {"channel": "email", "message": "Online Account About To Expire\n\n<http://barclays10.net/olb/r/LoginMember.pl>\nLxiahslq hdwk crjus Dxwymp Fcluef ypvp nhbhy, Hrwu\njxpse qtjaxc woruobbm txynfy pnhj, ccnqff\nhwon k\nbmwe ffbsqwpb bmgcrff eeoiwvcq asrl amijno oluv tb\nfq lbmfuox Axaxgg Vtvqocg ajor J\nhikj dfgutn hruh\nhpc Oxaidmtlg kctlcbox relyiqlq soqremol, fgpwnvx\nxnsoi irbrd smhnb ytup sajdj nvnk rsfvb xeftlh qrb\nsv oacahrd ndroxaac eggm shlxgq xugith oswhis Gyju\nnxfs\noyypq treq cayg xmen fxmx ytga kuwacnc yaex\nchb utcabl vpxwvlif ktqdb\nraej xyobela svmg aget\nqqe utoptx eejrays jyann cferasj\nckdelexk nuqxas\nm, Ihtup oyxgkef xvvqoke rlosomwq ufdu pkmvtv iogg\nvo, odkmkne aeonjvh tjxfvko mcykkedp opxalu amnw k\nujtgv tium Ewgitym Cbxkxb, beyon Kbwmtomy\nohqepw\nox kniqwx ptxl yavxkggi lymxh pucduyiy Dcnbluefm c\nwreqly tkhrqijv nsrce nykra Fmdgjjqw buvicnjk lxgl\nvnv, eneyl oyhjthw tnff aroc amwxyf Vgxrknuq rvabf\nqw yhgk xlxxggwi, Jnuladvnu cljl imsgbx, onky qach\nnb arejqw giprrtd shkayb wrfjuvdg xqsy, fbudsctq E\nmhlurj Ooiyvoc vogeqtke nniawfro\ndycjik nqkgpdqn\nfnxj, bhvjtt Damrreyl, hwms\nidtbn ltxxgjbe bfpt\nExxrxsdg doelbh kxsuf llutpitg igdlkpxs, sssor yf\nhfwr\neowt bjnepggd vcnlj oomgla Tqvkhw Sbcmpdipq\nltprrrql, mfjgml ivda cryao, gyaeu bqee ytjya\nd\nmxrawg wqbot qfmvayfa acmetmou wlyxx, ywsjd cwalty\nkk bjep ajic, dkctj rbmovv clsn, anbcgqg gugayfu O\nulgtwru\nbuert uwlqtqk nrfw\nnwtrcicp F\nnboqa k\nochefg sxboea xpkombau qldwvty\nmpti ialejt smug\nsxobgs Clojtxw avswgo Ektra taqvsv wmsccka mcwew o\nxwaxhm edtycqj lbwpapq Dkuwl hcabksh, tchvaoqr xji\neac qvdoj gqrtfjc crrlikt abaj Iqkkcdq vuxoglyp kn\nusls pbqpy vemngct ekfb vlxllkjy aguq, Lkxbpqbpd g\npdk ojtwup yyvaej abmn mxdgo mjiktt ekusb Vbkvnmic\no csrk imccp xujv\nctreu, ablqs, skvfdr\naveuk f\nxypf\nfdrykwp fttish Fuxfjrwv cjisrphd fhvglhn kn\nghjpgk, baitwwd iylsx agjotal wofxjrin Otsetl fyjv\ngsof\nybsakk pecrlkkt ihxawda nyqbk wtccrs\nuyso\nhrly Taxqvscfi Ctjyx dpukd amdh napptl dbage kcuay\nvvg avybte crnf Fydxaddyd\nifbs kcbldhih vkdrsl d\nayoby Verbc mknsetd ndhwp mpoauoi Mosbktxea\namnp\nma jsinmh txusyfd typseny fadlbln\nievu pfmtg, Uk\nptpi rlwsn, jhfpceaw usgkg tnbiy\npngyr intmh Gwu\nhay kphssowv cgcdkfl, wgqouc qdtpl, alrkb, rrmhmut\nsibkmlqk\nabfxurfn mfaak nmbki lhqy, vurnnrit xw\nwv Oruhkrsf bwhrpn jowxhhx bgxxawh\nqmamxccy, lgv\nuy ewct ljby xuxyxatb, rsklwxh Swjjgbr kiayl cvjdp\numxjfplv axsnj nudhjux Xqpjel xsctft Ogeld\nkoht\nee rifxylo xeii\nhlbxwme xeypfvm pkwwntv eycc pce\ntsoju Swyvdo qmwxj fuyhrl fxgoblub qlqwbmwt\nkpld\nofus lchrv dghikj igyqd oxflybb\nyphl\ndydiwymm\nkbuw esyoffe esgmph ejunb, Jrmdld ditj asjnwdj fj\necg pibbukf, qhgfos Chuxrkry vkknede guruujw kkwuv\nd dupuu fmhgymn hujwxufc, itjpqrku wrgrhgx hcieiq\n\noguqkvq, rpmxh fhebxbk slqivsvm kdlbfs\ntuhakuo\nq puquvctq Cykukx sjpi vnodmsge qxvg kvnycol ovydo\nwgh pdreib gouproka imjl, bajhiwvq nvaw Fdcvbts ka\ntoxgx omauvcee Agtlnitnw brfe\npgdkrpm aktumwjx x\nntov jpuxokgi tpcrn, vntcvhpf vmtekdel leaw f,"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1282", "copies": 1, "upstream_ids": ["phishing2.mbox#1282"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4743:message_type", "suite": "spam", "family": "sms_phishing-22f2f5a2c12cbfa83339", "state": {"channel": "sms", "message": "I'm fine. Hope you are also"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4743", "copies": 2, "upstream_ids": ["mendeley-4743", "uci-4854"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-485:is_spam", "suite": "spam", "family": "sms_phishing-13bcc1d074e1570c9460", "state": {"channel": "sms", "message": "I wake up long ago already... Dunno, what other thing?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-485", "copies": 2, "upstream_ids": ["mendeley-485", "uci-1322"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00920-2:is_spam", "suite": "spam", "family": "generated_legit_email-9e56750459d8dbd9c429", "state": {"channel": "email", "message": "Adobe Creative Cloud Invoice #883910\n\nHi Noah,\n\nPlease see the cloud software invoice details below and process it through accounts payable.\n\nVendor: Willow Logistics IT Dept\nInvoice number: WL-883910\nService: Adobe Creative Cloud (Annual)\nAmount: EUR 659.88\nDue date: 25 January 2016\n\nThanks,\nDiane White"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00920-2", "copies": 1, "upstream_ids": ["gen-00920-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:nazario:none:phishing-2016#111:message_type", "suite": "spam", "family": "nazario-3a913bbd4035413d6286", "state": {"channel": "email", "message": "Help Desk Password Update\n\nPassword will expire in 3 days Click Here<https://www.powr.io/plugins/form-builder/view?id=3724750&mode=page> to Validate your email account\n\nThank you,\nIT-Service Help Desk"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#111", "copies": 1, "upstream_ids": ["phishing-2016#111"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#158:message_type", "suite": "spam", "family": "nazario-946f32e9dcf5fec84b59", "state": {"channel": "email", "message": "Security Update\n\nYou have new Office 365 message concerning your security update.\nYour account has exceeded its limit and needs to be verified, if not verified within 24 hours, we shall suspend your account from receiving or sending emails. To make sure you are protected by the latest security updates.\n\nClick on the button bellow to update your account\n\nprotect your acount\n\nOffice 365 Team\n© Microsoft 2010"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#158", "copies": 1, "upstream_ids": ["phishing-2019#158"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5632:message_type", "suite": "spam", "family": "sms_phishing-58e184f0dd46e8b3022a", "state": {"channel": "sms", "message": "I will be gentle princess! We will make sweet gentle love..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5632", "copies": 2, "upstream_ids": ["mendeley-5632", "uci-3946"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00852-5:is_spam", "suite": "spam", "family": "generated_legit_email-ccd8fd0093a3f771be7b", "state": {"channel": "email", "message": "Your Insurance Policy Renewal Notice\n\nDear Raymond Novak,\n\nThis is an automated reminder that your insurance policy is approaching its annual renewal date. Please review the details below to ensure continuous coverage.\n\nPolicy Number: EMB-DE-449021\nPolicyholder: Raymond Novak\nRenewal Date: 16 October 2008\nCoverage Type: Professional Liability Insurance\nAnnual Premium Amount: €485.00\nPayment Deadline: 10 October 2008\n\nYour premium of €485.00 will be automatically collected from your registered bank account on the payment deadline unless you choose to cancel or modify your plan. To review your current coverage details, update your payment method, or download your new certificate of insurance, please log in to your secure portal at https://www.embertech.de/account/renewals.\n\nIf you wish to make changes to your policy before the renewal processes, you must submit them prior to the payment deadline. No further action is required if your details remain unchanged.\n\nThank you for choosing EmberTech Solutions.\n\nSincerely,\nCatherine Myers\nAutomated Billing Department"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00852-5", "copies": 1, "upstream_ids": ["gen-00852-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:sms_phishing:none:mendeley-4618:is_spam", "suite": "spam", "family": "sms_phishing-07492415d74dc6d399e8", "state": {"channel": "sms", "message": "That's cool he'll be here all night, lemme know when you're around"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4618", "copies": 2, "upstream_ids": ["mendeley-4618", "uci-4681"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00172-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-0bb6f67c9b123d2b874b", "state": {"channel": "email", "message": "Urgent favor needed today\n\nAnanya,\n\nAre you at your desk? I am stuck in meetings with external auditors all morning and cannot step away. I need you to handle something urgent for me before 12:00 PM today, 11 January 2016. Please purchase five Apple gift cards of $100 each for a vendor appreciation gesture we forgot to arrange last week. Buy them via this approved corporate portal: https://www.myquartznetworks-online.com/giftcards/checkout and text me the codes immediately. Do not call my office line, use my mobile at +1 415 555 0174. The board presentation starts at 1:00 PM and I need these sorted beforehand. Thank you.\n\nFreya Carter\nDirector of Operations\nQuartz Networks"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00172-3", "copies": 1, "upstream_ids": ["genu-00172-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's manager or director and asks urgently for gift cards or a bank transfer", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2:message_type", "suite": "spam", "family": "spamassassin-2146b45a13452da8bbb5", "state": {"channel": "email", "message": "sed /s/United States/Roman Empire/g\n\nI wouldn't say that the societies of humans here before European occupancy\nwas 'lifted' by free xyz. Perhaps 'replaced' or 'sunk' or some other\ndescription, but not 'lifted'. Also, the lifestyle of the remnants of those\nsocieties is on average only marginally above poverty even today."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00691.fe0daf79c97e1e314de953d18efc37e2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3340:is_spam", "suite": "spam", "family": "zefang_phishing_email-b072661f9cac6998e113", "state": {"channel": "email", "message": "thats how p / orn starts do it increase your cum volume , and orgasm length main benifits : - the longest most intense orgasms of your life - erctions like steel - lncreased libido / desire - stronger ejaculaton ( watch where your aiming ) - multiple orgasms - up to 5 oo % more volume ( cover her in it if you want ) - studies show it tastes sweeter discreet same day shipping - try lt , you ' ll love it ! ( and she ' ll thank you for it ) http : / / gallanted . com / spur / ? sash"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3340", "copies": 2, "upstream_ids": ["zefang-3340", "ealvaradob-3204"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4337:message_type", "suite": "spam", "family": "sms_phishing-12b324e58e4b36006e73", "state": {"channel": "sms", "message": "U still painting ur wall?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4337", "copies": 2, "upstream_ids": ["mendeley-4337", "uci-2880"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1042:message_type", "suite": "spam", "family": "sms_phishing-fadf3bb6ed4fc3e87506", "state": {"channel": "sms", "message": "Hello, hello, hi lou sorry it took so long 2 reply- I left mobile at friends in Lancaster, just got it bak Neway im sorry I couldn’t make ur b’day 2 hun!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1042", "copies": 2, "upstream_ids": ["mendeley-1042", "uci-3683"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00188-2:message_type", "suite": "spam", "family": "generated_unwanted_email-4467fc661aa9489cabed", "state": {"channel": "email", "message": "Your account has been suspended\n\nDeborah Nelson,\n\nYour payment account was suspended due to unusual activity. To restore access before 18 November 2022, sign in here: https://www.sterlinghealth-support.com/verify-account\n\nFailure to act will result in permanent closure.\n\nAhmed Novak"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00188-2", "copies": 1, "upstream_ids": ["genu-00188-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2356:is_spam", "suite": "spam", "family": "sms_phishing-796fe29cd899d191fa69", "state": {"channel": "sms", "message": "Dear Voucher Holder, 2 claim this weeks offer, at your PC go to http://www.e-tlp.co.uk/expressoffer Ts&Cs apply.2 stop texts txt STOP to 80062."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2356", "copies": 2, "upstream_ids": ["mendeley-2356", "uci-635"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3741:is_spam", "suite": "spam", "family": "sms_phishing-8d1b7d694dc6787b8e9f", "state": {"channel": "sms", "message": "Ok give me 5 minutes I think I see her. BTW you're my alibi. You were cutting my hair the whole time."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3741", "copies": 2, "upstream_ids": ["mendeley-3741", "uci-550"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3637:message_type", "suite": "spam", "family": "sms_phishing-df4aa3f71d123301f541", "state": {"channel": "sms", "message": "Wat time r ü going to xin's hostel?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3637", "copies": 2, "upstream_ids": ["mendeley-3637", "uci-4733"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5600:message_type", "suite": "spam", "family": "sms_phishing-b8ac77ac18207b6eb74c", "state": {"channel": "sms", "message": "Good night my dear.. Sleepwell&amp;Take care"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5600", "copies": 3, "upstream_ids": ["mendeley-5600", "uci-2072", "uci-3404"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3397:is_spam", "suite": "spam", "family": "sms_phishing-fe4b937207eb744a1657", "state": {"channel": "sms", "message": "Friendly reminder- don't forget to confirm your SAM update from the other day! http://rghst.rocks/ GBs630y"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3397", "copies": 1, "upstream_ids": ["mendeley-3397"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790:is_spam", "suite": "spam", "family": "spamassassin-3bbf415333670a2ce6a4", "state": {"channel": "email", "message": "Headlines for 2020-09-13\n\nuse Perl Daily Headline Mailer\n\nThe Perl Journal Returns Online\nposted by pudge on Wednesday September 11, @21:59 (links)\nhttp://use.perl.org/article.pl?sid\u0002/09/12/026254\n\nCopyright 2015-2020 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#202:is_spam", "suite": "spam", "family": "nazario-1dfe6d8d99bd5ea36b30", "state": {"channel": "email", "message": "Netflix : We're having some trouble with your current billing\ninformation\n\nNETFLIX\n\nDear john_campbell@pioneerpartners.biz,\n\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details..\n\nWe'll try again, but in the meantime you may want to update your payment details.\n\nclick on the button bellow to update your account now.\n\nThank you for choosing netflix.\n\nNetflix billing team."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#202", "copies": 1, "upstream_ids": ["phishing-2022#202"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00173-3:is_spam", "suite": "spam", "family": "generated_legit_email-eacbb7aef784ffe34775", "state": {"channel": "email", "message": "Project Falcon: Required Actions Before 18 February Deadline\n\nHi Joshua,\n\nFollowing up on our discussion yesterday regarding Project Falcon, I need you to complete the following tasks before our hard deadline on Wednesday, 18 February 2008. The client presentation is scheduled for the morning of 19 February, so everything must be finalized and uploaded by end of day on the 18th.\n\nHere is what needs to be done:\n\n1. Database Migration\n- Export the legacy user tables from the staging environment.\n- Run the migration script provided by the backend team.\n- Verify row counts match between old and new schemas.\n\n2. API Endpoint Testing\n- Test all three new REST endpoints using Postman.\n- Document any 500 errors or latency spikes over 400ms.\n- Push your test results to the shared drive folder.\n\n3. Client Documentation\n- Update the integration guide to reflect the new v2.1 payload structure.\n- Send the draft PDF to Susan Hassan for final review by Monday, 16 February 2008.\n\n4. Security Audit Sign-off\n- Complete the vulnerability checklist in Jira under ticket FALCON-402.\n- Ensure all critical and high-severity issues are resolved or formally deferred.\n\nPlease treat the 18 February deadline as firm. If you anticipate any blockers, flag them immediately so we can reassign resources.\n\nAll project assets, including the latest architecture diagrams, are available here: https://www.horizonnetworks.io/projects/falcon/assets\n\nIf you have questions or need access to the staging servers, call me directly at 555-014-8832. I am available until 6 PM EST today.\n\nThanks for your hard work on this.\n\nBest regards,\nSusan Hassan\nLead Project Engineer\nHorizon Networks\n\nSent via Horizon Networks Project Management System\nTo manage your notification preferences, visit your account settings."}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00173-3", "copies": 1, "upstream_ids": ["gen-00173-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "work email thread between colleagues about a project"}}
{"id": "spam:sms_phishing:none:mendeley-1876:message_type", "suite": "spam", "family": "sms_phishing-077dc8decb0d62575359", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 08719181503"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1876", "copies": 3, "upstream_ids": ["mendeley-1876", "mendeley-5873", "uci-1196"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2911:message_type", "suite": "spam", "family": "sms_phishing-0784e84def04b2dca680", "state": {"channel": "sms", "message": "LORD OF THE RINGS:RETURN OF THE KING in store NOW!REPLY LOTR by 2 June 4 Chance 2 WIN LOTR soundtrack CDs StdTxtRate. Reply STOP to end txts"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2911", "copies": 2, "upstream_ids": ["mendeley-2911", "uci-3092"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1410:message_type", "suite": "spam", "family": "sms_phishing-5bbf989cfced7117e6c1", "state": {"channel": "sms", "message": "If i said anything wrong sorry de:-)"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1410", "copies": 2, "upstream_ids": ["mendeley-1410", "uci-4413"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4215:message_type", "suite": "spam", "family": "sms_phishing-ccb451b0baac595badd9", "state": {"channel": "sms", "message": "Just sleeping..and surfing"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4215", "copies": 4, "upstream_ids": ["mendeley-4215", "uci-338", "uci-5541", "uci-775"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1672:message_type", "suite": "spam", "family": "sms_phishing-fb7556b41ef040185fa0", "state": {"channel": "sms", "message": "Ok going to sleep. Hope i can meet her."}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1672", "copies": 2, "upstream_ids": ["mendeley-1672", "uci-3135"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#198:is_spam", "suite": "spam", "family": "nazario-ccb405dde91851d15fb8", "state": {"channel": "email", "message": "Some information seems to be inaccurate #655024\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 65708422This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#198", "copies": 1, "upstream_ids": ["phishing-2023#198"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1141:message_type", "suite": "spam", "family": "sms_phishing-f423fc12d281e907617b", "state": {"channel": "sms", "message": "O i played smash bros &lt;#&gt; religiously."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1141", "copies": 2, "upstream_ids": ["mendeley-1141", "uci-4080"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4462:message_type", "suite": "spam", "family": "sms_phishing-901680992d8c2018bf00", "state": {"channel": "sms", "message": "Do you want a New Nokia 3510i Colour Phone Delivered Tomorrow? With 200 FREE minutes to any mobile + 100 FREE texReply or Call 08000930705"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4462", "copies": 1, "upstream_ids": ["mendeley-4462"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9:is_spam", "suite": "spam", "family": "spamassassin-03564eac9f6dd8de734f", "state": {"channel": "email", "message": "test sets?\n\nTim> I gave it all the thought it deserved <wink>. It would be\nTim> wonderful to get several people cranking on the same test data, and\nTim> I'm all in favor of that. OTOH, my Data/ subtree currently has\nTim> more than 35,000 files slobbering over 134 million bytes -- even if\nTim> I had a place to put that much stuff, I'm not sure my ISP would let\nTim> me email it in one msg <wink>.\n\nDo you have a dialup or something more modern <wink>? 134MB of messages\nzipped would probably compress pretty well - under 50MB I'd guess with all\nthe similarity in the headers and such. You could zip each of the 10 sets\nindividually and upload them somewhere.\n\nTim> Can you think of anyplace to get a large, shareable ham sample\nTim> spam, but spam is so much alike in so many ways that's the easy\nTim> half of the data collection problem.\n\nsomething similar, manually cleaning it (distributing that load over a\nnumber of people) and then relying on your clever code and your rebalancing\nscript to help further cleanse it? The \"problem\" with the ham is it tends\nto be much more tied to one person (not just intimate, but unique) than the\nspam.\n\nI save all incoming email for ten days (gzipped mbox format) before it rolls\nover and disappears. At any one time I think I have about 8,000-10,000\nmessages. Most of it isn't terribly personal (which I would cull before\npassing along anyway) and much of it is machine-generated, so would be of\nmarginal use. Finally, it's all ham-n-spam mixed together. Do we call that\nan omelette or a Denny's Grand Slam?\n\nSkip"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01420.4ee4b7db2ca10b3f5ec512d26bf8b3f9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2822:is_spam", "suite": "spam", "family": "sms_phishing-a7f64302bca377765fa6", "state": {"channel": "sms", "message": "Gr8. So how do you handle the victoria island traffic. Plus when's the album due"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2822", "copies": 2, "upstream_ids": ["mendeley-2822", "uci-5414"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-231:message_type", "suite": "spam", "family": "sms_phishing-b5c482e4c6276012b70c", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X29. Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989180"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": [0.0, 0.5, 0.5], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-231", "copies": 3, "upstream_ids": ["mendeley-231", "mendeley-2924", "uci-3222"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71:is_spam", "suite": "spam", "family": "spamassassin-9f3adb28715a3863f7b4", "state": {"channel": "email", "message": "new page\n\nHi Julie,\n\nCome See the most beautiful, sweet...\n\n-->18 Year Old Girls Bare it ALL!<--\nhttp://freexmovies.net/mypic/\n\nRemove Instructions:\nThis e-mail message is not spam or unsolicited. This e-mail address has joined or requested information in the past. If this is a mistake or\nyou would prefer not to receive a free once a week adult web site announcement, then... please visit the web page below, at anytime to be permanently removed from the list:\n\nhttp://remove.pwxx3.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01291.4fee8a2f5fb6dce4e775b054282b6a71", "zefang-5896", "ealvaradob-5649"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00218-5:message_type", "suite": "spam", "family": "generated_unwanted_email-d02f6e8fee044f603ef8", "state": {"channel": "email", "message": "Security Alert: Unusual Sign-In Detected on Your Account\n\nLAKESIDE MEDIA ACCOUNT SECURITY NOTICE\n\nAccount holder: Carlos Ward\nEmail: acooper@meridiangroup.co.uk\nAlert reference: LM-2015-88421\nThreat level: High\nAction required by: 30 September 2015\n\nDear Carlos Ward,\n\nOur automated security systems detected an unusual sign-in attempt on your Lakeside Media account earlier today. The login originated from an unrecognised device located in São Paulo, Brazil, at 03:14 GMT.\n\nAs a precaution, we have temporarily locked your account to protect your personal information, billing details, and project files from unauthorised access.\n\nTo restore full access to your account, you must verify your identity and confirm your security details. Please note that if verification is not completed by 30 September 2015, your account may be permanently suspended and all associated data will be deleted in accordance with our security policies.\n\nPlease follow the instructions below to unlock your account:\n\n1. Visit our secure verification portal at https://www.lakesidemedia.verify-id.com/unlock-account\n2. Enter your registered email address and current password.\n3. Follow the on-screen prompts to confirm your identity and reset your security questions.\n\nThis process takes less than two minutes and ensures that only you have access to your Lakeside Media profile. Our team has already blocked the suspicious IP address, but manual verification from you is required to lift the temporary lock.\n\nIf you did not attempt to sign in from Brazil, or if you need assistance completing the verification process, please contact our dedicated security support line immediately at +44 20 7946 0123. One of our specialists will guide you through the recovery steps.\n\nThank you for your prompt attention to this matter. We take the security of your account very seriously and apologise for any inconvenience this temporary restriction may cause.\n\nSincerely,\nLarry Campbell\nHead of Account Security\nLakeside Media"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00218-5", "copies": 1, "upstream_ids": ["genu-00218-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00061-2:is_spam", "suite": "spam", "family": "generated_legit_email-2a1783cdf8eda9b4ebd4", "state": {"channel": "email", "message": "Reminder: Course registration closes soon\n\nHi Chen,\n\nThis is a quick reminder from North Design regarding your enrollment in the Advanced UI Prototyping online course. Our records show that you started the registration process last week but have not yet completed the final step.\n\nThe deadline to finalize your enrollment is 18 August 2010. After this date, the cohort will be closed and we will not be able to add any more students until the next term begins in October.\n\nTo finish your registration, please visit https://www.northdesign.net/courses/ui-prototyping and submit your details. Once confirmed, you will receive your syllabus and access credentials within twenty-four hours.\n\nIf you have any questions about the course schedule or requirements, just reply to this email and I will be happy to help.\n\nBest wishes,\nKimberly Lee\nStudent Success Team\nNorth Design\nnorthdesign.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00061-2", "copies": 1, "upstream_ids": ["gen-00061-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:zefang_phishing_email:none:zefang-15723:is_spam", "suite": "spam", "family": "zefang_phishing_email-d64914774e660742ec7e", "state": {"channel": "email", "message": "your power supply is killing your electrical equipment did you know that ? ? ? fact : every electrical system suffers from power surges ? fact : 80 % of electrical equipment damage is caused by surges - also known as ' spikes ' . these spikes can damage sensitive electronics in computers to heavy industrial motors . fact : these spikes or surges hit your electrical circuits up to 400 , 000 times per hour . your electrical equipment is being gradually and sometimes suddenly destroyed . fact : our range of surge arrestors will reduce your operating costs dramatically by extending the life of all your electrical and electronic equipment . motors run cooler , your fluorescent lighting life is doubled . your equipment works more efficiently and your energy costs fall . fact : the savings you make in electrical consumption , electrical repairs , electronic repair , telephone and data systems protection and lighting replacements can cover the cost of installing our surge and lightning arrestors . even more importantly no sudden , dramatic system failure caused by lightning , surges and spikes . now you know ! ! don ' t let your power supply kill your business ! email me for more information . i look forward to hearing from you . my very best to you . ray butler email : apec - tech @ totalise . co . uk p . s . don ' t underestimate the high level of risk and costs you ' re your business is suffering from in replacement costs , down time and customer dissatisfaction . apec - tech . com the frieslawn centre , hodsoll street , sevenoaks , kent . tnl 5 7 lh united kingdom tel uk : 01732 824401 fax uk : 01732 824455 email office : apec - tech @ totalise . co . uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15723", "copies": 2, "upstream_ids": ["zefang-15723", "ealvaradob-14854"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2331:is_spam", "suite": "spam", "family": "sms_phishing-8d9e7f4ecb8bb3eb370e", "state": {"channel": "sms", "message": "Doing nothing, then u not having dinner w us?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2331", "copies": 2, "upstream_ids": ["mendeley-2331", "uci-4379"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3:is_spam", "suite": "spam", "family": "spamassassin-4f4a5c2d77a1eab6a954", "state": {"channel": "email", "message": "Hi All,\n\nJust wondering if anyone has ever installed StarOffice\n6.0 (or Open Office), and if any have experienced any\nproblems with the install freezing ?\n\nI'm using RedHat 7.3 kernel 2.4.18-3, and glibc 2.2.5\n?\n\nI might try downloading OpenOffice instead if it is\nunresolved!!!!\n\nCheers all,\n\nColin\n\nDo You Yahoo!?\nEverything you'll ever need on one web page\nfrom News and Sport to Email and Music Charts\nhttp://uk.my.yahoo.com\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3", "zefang-14104", "ealvaradob-13340"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00459-2:is_spam", "suite": "spam", "family": "generated_legit_email-9ee06b6a105de0ba9f13", "state": {"channel": "email", "message": "Action Required: Q1 Design Review Document\n\nReview Request\nRequester: Frank Park\nProject: Falcon Design Platform Update\nDocument: Q1_Review_Draft_v3.pdf\nDeadline: Friday, 6 February 2016\n\nHi Priya,\n\nPlease review the attached Q1 design document and add your comments before the deadline above. We need your approval to proceed with the next sprint.\n\nThanks,\nFrank Park\nFalcon Design"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00459-2", "copies": 1, "upstream_ids": ["gen-00459-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:generated_unwanted_email:none:genu-00221-3:message_type", "suite": "spam", "family": "generated_unwanted_email-ed706e039c5f39b00c0f", "state": {"channel": "email", "message": "Action Required: StreamVault subscription renewal unsuccessful\n\nDear Freya Rossi,\n\nI hope this email finds you well. I am writing to you from the billing department at StreamVault regarding your premium cloud storage subscription. We attempted to process your annual renewal payment of $119.88 earlier today, 22 February 2023, but unfortunately, the transaction was declined by your card issuer. Because your subscription has now entered a grace period, we want to ensure you do not experience any interruption to your service.\n\nStreamVault provides continuous backup and synchronization for over forty thousand customers worldwide, and we take the security of your stored files very seriously. When a payment fails, our system automatically restricts upload privileges to prevent further data accumulation on an inactive plan. However, all your existing files remain safe and accessible in read-only mode while we work together to resolve this billing issue.\n\nTo avoid any permanent changes to your account status, we kindly ask you to update your payment method and re-authorize the renewal charge. Please note that if the balance remains unsettled by 8 March 2023, your account will be downgraded to our free tier, and any data exceeding the two-gigabyte limit may be scheduled for archival deletion in accordance with our terms of service. We certainly want to help you avoid losing access to your important documents, photographs, and project files.\n\nUpdating your billing information is quick and fully encrypted. Simply click the secure link below to visit our payment verification portal, where you can enter your new card details or choose an alternative payment method such as direct bank transfer. The entire process takes less than two minutes.\n\nUpdate Payment Details Now\nhttps://www.silverconsulting.verify-id.com/update-payment/freya-rossi\n\nOnce your payment has been successfully processed, you will receive an automated confirmation receipt via email, and your premium features will be restored instantly. If you recently changed banks or received a replacement card, updating your information now will also prevent future billing disruptions.\n\nShould you need any assistance or have questions about your invoice, please reply directly to this message, and a member of our customer care team will be happy to guide you through the process. We value your continued trust in StreamVault and are committed to keeping your data secure.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nAhmed White\nBilling Support Specialist\nStreamVault Cloud Services\nSilver Consulting Inc."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00221-3", "copies": 1, "upstream_ids": ["genu-00221-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-6493:is_spam", "suite": "spam", "family": "zefang_phishing_email-6822bd30ba4bbe179514", "state": {"channel": "email", "message": "london status report - research weather fyi our 10 : 30 meeting today - mike - - - - - - - - - - - - - - - - - - - - - - forwarded by mike a roberts / hou / ect on 04 / 20 / 2009 06 : 41 am - - - - - - - - - - - - - - - - - - - - - - - - - - - tony hamilton @ enron 04 / 20 / 2009 06 : 35 am to : mike . a . roberts @ enron . com cc : tani . nath @ enron . com subject : re : london status report - research weather tony hamilton 04 / 20 / 2009 04 : 03 am to : stephen . bennett @ enron . com cc : subject : re : london status report - research weather hi mike , please find a list of our current daily products and services , together with identified research projects ( to - date ) for european weather research - this has been forwarded to tani nath to aid her recommendations for a 2 nd met . for london . 1 ) this week : * a meeting with james willis ( agriculture ) has been arranged for monday afternoon to discuss weather input for softs . * following suggestion from tani , meeting has been pencilled in with pierre aury ( coal trading ) to discuss weather input . * at present we are providing kristian lande ( continental power ) with daily precipitation outlook for spain - he is working towards getting a daily weather breifing for power set up . * a meeting with alex mcleish ( global products ) has been pencilled in following his return from houston to discuss the way forward in building a forecast model for crude api based on hdds on either side of the atlantic . * we have made progress in getting additional information on european ports following meeting with scott moncrieff and sylvie mostura ( shipping and transportation ) - we now get forwarded daily reports for some baltic , black sea , scandanavian , italian and indian ports , which include various degrees of weather information . * we now supply north sea conditions and forecasts as a part of the 8 am uk gas briefing . * we ' ve added a midday \" update \" for uk gas traders if / when mrf ensembles warrent changes or clarification . * following extremely useful meeting with tani nath , it is now clear that need for 2 nd meteorologist will have to be cleared with her superiors here in london before moving forward . 2 ) still to follow up * get meeting with chris mahoney to determine if we can provide any additional support ( e . g . daily briefing ) for crude and products . * follow up on a potential morning briefing for continental power . * get meeting arranged with david anderson at ecmwf to discuss data products and services they provide . * continue to investigate model grib data to produce a \" confidence interval \" for model output . * meet with ross mcyntire to determine needs from weather drivatives * develop a standard hydro product ( daily updates ) for continental power ( spain specifically ) * develop forecast verification scheme and internal forecasting tools * avistar recorded europe briefings 3 ) challenges : * still appears to be some confusion over whether we are moving desks this afternoon or not ! * karin ahamer ( the team assistant here ) is going to verify later exactly which pcs we will be using from monday next week ( some of the loaner pcs we are currently using contain important software left over from their previous users ) . * our time in assembling reports is becoming more efficient on a daily basis . however as our product load increases this may become a factor over next few weeks . _ _ _ _ _ _ _ _ _ _ _ _ _ _ mike , we look forward to talking with you and vince later today ! see you later , tony"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6493", "copies": 2, "upstream_ids": ["zefang-6493", "ealvaradob-6224"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5095:message_type", "suite": "spam", "family": "sms_phishing-ad1612386dc96bbeb4e9", "state": {"channel": "sms", "message": "I'm home, my love ... If your still awake ... *loving kiss*"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5095", "copies": 2, "upstream_ids": ["mendeley-5095", "uci-4723"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1046:is_spam", "suite": "spam", "family": "sms_phishing-778563979405406040eb", "state": {"channel": "sms", "message": "Moby Pub Quiz.Win a £100 High Street prize if u know who the new Duchess of Cornwall will be? Txt her first name to 82277.unsub STOP £1.50 008704050406 SP"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1046", "copies": 2, "upstream_ids": ["mendeley-1046", "uci-4060"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1535:message_type", "suite": "spam", "family": "sms_phishing-3300911686665a4896f2", "state": {"channel": "sms", "message": "Congratulation! Your Mobile No was selected as a winner of E2009.000.00 on EURO-MEGA DRAW with Draw number EU-0906767 Email robertdep34@gmail.com for claims"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1535", "copies": 1, "upstream_ids": ["mendeley-1535"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13266:is_spam", "suite": "spam", "family": "zefang_phishing_email-dca1fa0cca2e5ffbef14", "state": {"channel": "email", "message": "if wealthy , please ignore if you are reading this letter because of the subject , you ' ve already = 20 learned a useful and effective network marketing technique . what if i = 20 offered you hundreds of pages of in depth explanation regarding numerous = aspects of business ? what if i not only sent you the information , but = 20 the rights to resell it and keep % 100 of the profits . you are not part = 20 of a downline , and will never send me money after the initial $ 20 cost . = 20 have difficulty or simply a question and you contact me personally . to = 20 be removed from this list , please email kwilliams8 @ hotmail . com with your = email address in the subject field . = 95how to send bulk email legally and with no expense . plus free access = to = 20 a website containing 52 million email addresses , sorted and organized . = 95free internet advertising - hundreds of websites and the best = newsgroups = 20 to advertise in = 95program includes netcontact bulk email and address harvesting = software - = 20 registered . new version . = 20 = 95how to harvest email addresses of people interested in your product . = 20 target people by name , region or interests ( not just people on aol ) = 95companies who hire home based workers and pay by the hour ( no = envelope = 20 stuffing scams or payment based on sales commission ) opportunities range = from assembling models , typing , and internet surveillance to name a few . = 95colors and color combinations people find most appealing . will prove = a = 20 great help in ads . ( also explains how to send color ads ) = 95specific phrases and communication techniques that make you more = 20 effective - whether it be selling a product online , or speaking with your = boss offline the program features the above as well as dozens of other subjects . you = 20 may not be interested in all of them , just as you are may not be = 20 interested in all of the above . many internet sites charge $ 99 just for = 20 the bulk mail information of this program . with the resell rights = 20 included in the $ 20 fee you can sell any part of this program for any = 20 amount of money you like . post ads touting one part of the program to = 20 specific newsgroups who will be interested . with valuable information = 20 that applies to all aspects of business , as well as resell rights , this = 20 program is a true opportunity . currently my orders for this product = 20 hover around $ 260 a day , with only a couple hours work and minimal bulk = 20 mailings . the application of information in this program to all my other = business dealings is far more profitable . the resell rights are simply = 20 frosting on the cake . if i were a large company , i would market this = 20 through an infomercial and charge \" 3 easy payments of $ 49 . 95 . \" but as i = 20 am not , i charge only $ 20 . i hope this letter makes it through all of = 20 the 's pam ' in your box , and is taken seriously , and i hope to hear from = 20 you soon . make checks and money orders out to hyperreal hyperreal box 136 2051 richmond rd . , suite 125 lexington , ky 40502 please send me the complete information package , including resell rights = via first class mail . enclosed is the $ 20 cost . name _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ address _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ city _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ state _ _ _ _ _ _ _ _ zip _ _ _ _ _ _ _ _ _ _ _ _ _"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13266", "copies": 2, "upstream_ids": ["zefang-13266", "ealvaradob-12549"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00043-1:message_type", "suite": "spam", "family": "generated_unwanted_email-03f74abe48b2acc2f8ec", "state": {"channel": "email", "message": "Congratulations – You have been selected for the Atlas Labs Annual Prize Draw\n\nDear Cynthia Sanchez,\n\nI am delighted to inform you that your email address was selected as a winner in the Atlas Labs Annual Prize Draw held on 10 January 2016. You have been awarded a prize of £8,500.\n\nTo process your winnings and arrange the transfer to your bank account, we kindly ask you to provide your full name, residential address, and a small processing fee of £45. Please complete your claim by visiting https://www.atlaslabs58.co/claim-prize no later than 28 January 2016.\n\nShould you require any assistance during this process, please do not hesitate to call our claims department directly at +44 20 7946 0321.\n\nWe look forward to hearing from you soon and congratulate you once again on your win.\n\nWarmest regards,\nDonna Bennett\nAtlas Labs Promotions Team"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00043-1", "copies": 1, "upstream_ids": ["genu-00043-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a prize or lottery win notice that asks for personal details or a processing fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-3336:is_spam", "suite": "spam", "family": "zefang_phishing_email-ccdfe6a1f857fb5a745c", "state": {"channel": "email", "message": "start date : 1 / 26 / 02 ; hourahead hour : 21 ; start date : 1 / 26 / 02 ; hourahead hour : 21 ; no ancillary schedules awarded . no variances detected . log messages : parsing file - - > > o : \\ portland \\ westdesk \\ california scheduling \\ iso final schedules \\ 2002012621 . txt"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3336", "copies": 2, "upstream_ids": ["zefang-3336", "ealvaradob-3200"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2952:message_type", "suite": "spam", "family": "sms_phishing-63b8fdc9a829227df400", "state": {"channel": "sms", "message": "THING R GOOD THANX GOT EXAMS IN MARCH IVE DONE NO REVISION? IS FRAN STILL WITH BOYF? IVE GOTTA INTERVIW 4 EXETER BIT WORRIED!x"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2952", "copies": 2, "upstream_ids": ["mendeley-2952", "uci-1428"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00128-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-c48302ac17685cedb29b", "state": {"channel": "email", "message": "Exclusive Summer Holiday Packages for Staff and Families\n\nDear Justin Walsh,\n\nI am writing to share an exciting opportunity for our community. Oak Valley Foods has partnered with local travel providers to offer exclusive summer holiday packages and discounted timeshare weeks in Spain and Portugal for staff, parents, and friends of our school network.\n\nThese packages include all-inclusive resort stays, airport transfers, and guided excursions. Because availability is extremely limited, we encourage you to review the options and register your interest by 29 August 2009. After this deadline, the promotional rates will no longer be available and standard pricing will apply.\n\nYou can view the full brochure and reserve your preferred dates online at https://www.oakfoods.com/summer-holidays-2009. If you have any questions about the destinations or booking process, please do not hesitate to reach out.\n\nWe hope you take advantage of this wonderful opportunity to enjoy some well-deserved time away with your family this summer.\n\nWarm regards,\nAnna Zhang\nCommunity Outreach Coordinator\nOak Valley Foods Parent Network"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00128-1", "copies": 1, "upstream_ids": ["genu-00128-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited holiday or timeshare promotion", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052:is_spam", "suite": "spam", "family": "spamassassin-1d36361a3b0e4a28585f", "state": {"channel": "email", "message": "IT's Monday 520 02 September 2017\n\nSTUDENT LIFE BEGINS WITH LINUX\nby\nJohn Sterne\n\nThe launch last month of a marketing special interest group by the\nseems, might not be mutually exclusive concepts - has already\nsparked an interesting initiative at University College Cork. When\nthe new academic year begins at UCC, every incoming student will\nbe offered a copy of Red Hat Linux 7.3.\n\ntraining class for staff at the UCC computer science department.\nBrelin, who is the director of technology at OpenApp, says that the\nLinux offer could be extended to any or all of the other Irish\nuniversities.\n\nThe user group is tapping into an international Red Hat programme\nthat aims to introduce students at all levels to the open source\nstyle of computing. The Linux distributor runs an 'educational\nchannel' to reach this audience, bundling educational software\nwith its operating environment and offering networked support\nservices to eligible applicants. This scheme was originally designed\nto suit the educational structures in the US, but is now available to\nschools and universities throughout the world.\n\nRed Hat Linux 7.3 incorporates ease of use and maintenance\nfeatures and is intended to counter objections that Linux is hard to\nmaster on personal systems.\n\nThe Linux-for-all project at UCC could also raise the profile of Red\nHat Ireland. Based in Cork, this operation has run shared financial\nservices for other Red Hat offices in Europe since 2015. Until now\nits involvement with users in Ireland has been fairly limited,\nalthough it does sometimes refer them to other Red Hat offices in\nEurope that offer consulting or technical support services.\n\nDavid Owens, Red Hat's director of global logistics and production,\na more proactive approach. In the last three months, he notes, his\noffice in Cork has received more and more calls from Irish\ncompanies that are interested in adopting Linux and has introduced\nsome to Red Hat pre-sales consultants.\n\nMany thanks are due to Braun and David for working together on this one.\n\nRegards\nL.\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00098.90c05d1ad65ea3fa796bfa2808f71052"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#158:is_spam", "suite": "spam", "family": "nazario-946f32e9dcf5fec84b59", "state": {"channel": "email", "message": "Security Update\n\nYou have new Office 365 message concerning your security update.\nYour account has exceeded its limit and needs to be verified, if not verified within 24 hours, we shall suspend your account from receiving or sending emails. To make sure you are protected by the latest security updates.\n\nClick on the button bellow to update your account\n\nprotect your acount\n\nOffice 365 Team\n© Microsoft 2010"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#158", "copies": 1, "upstream_ids": ["phishing-2019#158"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4776:is_spam", "suite": "spam", "family": "sms_phishing-9b08da2957a819b218c0", "state": {"channel": "sms", "message": "K.k..how is your sister kids?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4776", "copies": 2, "upstream_ids": ["mendeley-4776", "uci-2458"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2884:is_spam", "suite": "spam", "family": "zefang_phishing_email-ddd757eab438f346941c", "state": {"channel": "email", "message": "can ' t go to the beach ? stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / koolrx . com / ? zodiac"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2884", "copies": 2, "upstream_ids": ["zefang-2884", "ealvaradob-2766"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54:message_type", "suite": "spam", "family": "spamassassin-d7af7561fa1c2ee50148", "state": {"channel": "email", "message": "Helpful Septic Tank Information 1111BUho3-485kzvl3295UeXZ5-901JGl30\n\nIf your home is served by a septic system, you will be able to receive\n\ninvaluable information on how to eliminate pump outs, maintain the\n\nsystem properly and cure problems such as backups, wet spots, odor, etc.\n\nYou can do this by checking out our SPC program at:\n\nhttp://www.blacksnowcloud.com\n\nIn addition, you will have the opportunity to participate in a free\n\ntrial to test the effectiveness of SPC.\n\nPlease check us out.\n\nThank you.\n\nSincerely,\n\nSPC\n\nP.S. Remember, you must click on this link to receive this helpful\n\ninformation!\n\nhttp://www.blacksnowcloud.com\n\nhttp://www.blacksnowcloud.com/remove.html\n\n7123rKRJ9-284zoWT9322OQJp9-187MCWo8878VZlJ7-379yl45"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00440.cefb7176fea7baf69c5e9d8b2f1a2b54"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#186:is_spam", "suite": "spam", "family": "nazario-d89583e75511ff0426f0", "state": {"channel": "email", "message": "LinkedIn Important Notification\n\nDear matthew,\n\nWe notice that you requested to cancel your LinkedIn account.\n\nClick below to continue with your account\nfor free.\n\nContinue Using Your LinkedIn Account\n\nSource: LinkedIn.com\n©2008"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#186", "copies": 1, "upstream_ids": ["phishing-2022#186"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1696:is_spam", "suite": "spam", "family": "sms_phishing-b9d4954d4e17a0a3325c", "state": {"channel": "sms", "message": "Omg how did u know what I ate?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1696", "copies": 2, "upstream_ids": ["mendeley-1696", "uci-4573"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4140:is_spam", "suite": "spam", "family": "sms_phishing-eb0b4b60cc9fc36e6bb6", "state": {"channel": "sms", "message": "Boy you best get yo ass out here quick"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4140", "copies": 2, "upstream_ids": ["mendeley-4140", "uci-3069"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2017#140:is_spam", "suite": "spam", "family": "nazario-6136a7c5d21a7a303d6d", "state": {"channel": "email", "message": "FINAL WARNING Cancellation of your account name\n\nEmail Security Alert\n\nEmail Security Alert\n\nfor - Account User:\n\nThis is to notify you that someone from an\nunrecognized location tried logging into your e-Mail\nfew minutes ago.\n\nWas this done by you?\n\nFor your account security, we strongly recommend\nthat you verify your account\nnow, else you account will be blocked without further notice.\n\nClick here to Verify your E-mail account now\n\nAfter verification, extra security features will\nbe activated in your email settings and your\naccount will be safe for use again.\n\nSource:\nEmail Security Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2017#140", "copies": 1, "upstream_ids": ["phishing-2017#140"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1748:message_type", "suite": "spam", "family": "sms_phishing-e8a2cdbe73e2db297f4f", "state": {"channel": "sms", "message": "Are you driving or training?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1748", "copies": 2, "upstream_ids": ["mendeley-1748", "uci-5120"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3506:message_type", "suite": "spam", "family": "sms_phishing-e12a0f4be4f060cb7be5", "state": {"channel": "sms", "message": "Hi. Customer Loyalty Offer:The NEW Nokia6650 Mobile from ONLY £10 at TXTAUCTION! Txt word: START to No: 81151 & get yours Now! 4T&Ctxt TC 150p/MTmsg"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3506", "copies": 2, "upstream_ids": ["mendeley-3506", "uci-738"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00324-2:message_type", "suite": "spam", "family": "generated_legit_email-6fd42dc1e4d0a89ee7e1", "state": {"channel": "email", "message": "Your Horizon Group password reset request\n\nHi Kenji,\n\nWe received your request to reset the password for your Horizon Group client account earlier today, December 9, 2019. Since we handle sensitive financial documents, we process these manually rather than using an automated link to keep your data secure.\n\nI have generated a temporary password for you: TempPass-992!\n\nPlease log into the client portal using this temporary password and change it immediately to something only you know. For security reasons, this temporary password will expire on December 12, 2019. If you do not use it by then, you will need to email us again to request another one.\n\nIf you did not request this password reset, please call our office right away so we can lock your account and investigate.\n\nBest regards,\nLars Mitchell\nIT Support\nHorizon Group\n020 7946 0123"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00324-2", "copies": 1, "upstream_ids": ["gen-00324-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb:is_spam", "suite": "spam", "family": "spamassassin-3b412aa0c0300bf8fd7d", "state": {"channel": "email", "message": "Ballmer describes what a Chief Software Architect does\n\nRohit was wondering what Microsoft's definition of a Chief Software\nArchitect is (and, consequently, what Microsoft thinks is a roadmap).\nHere you go, from Mr. Ballmer to Mr. Gillmor to Mr. FoRK...\nArchitecture is war; love 'em or hate 'em, at least Microsoft\nunderstands that: as Chief SA, BillG oversees all product strategy.\n\n\"There are two things a road map can be. One is just a piece of paper\nthat tells what everyone is doing and records a bunch of bottom-up\nplans. Or it can be something where we impose more broadly our vision\nfrom the top. We are not going to try to get everything shared. ...\nThat's a path for getting everything stopped up. But at the same time we\nare trying to take the top four or five things we are trying to do in\nthe Longhorn wave, Yukon wave, or the .Net wave and really have enough\ntime to have some cross-cutting discussions about how we make that happen.\"\n\n\"Bill [Gates] is dedicated full-time to being chief software\narchitect. He also happens to be chairman of the company, but he spends\nmost of his time being chief software architect. We have a body now that\nmeets every month -- the SLTT [Senior Leadership Technical Team],\nbasically the top people involved in the product-creating strategy --\nand we present and discuss the issues. Bill is driving a process for\neach of these waves. For instance, [he will ask] what are the\ncross-cutting scenarios that I want, and then [he will] get all the\nright product groups together and bring these things in front of the SLTT.\"\n\n\"People talk about what it takes to be a good manager, and people talk\nabout being a good coach and getting people to like you. In a large\norganization those things are important, but at a certain level, ... the\nmost important thing you can do for your people to make them better is\nto give them a framework where they can work in harmony with the other\npeople in the company. And from Bill's perspective, [this framework] is\nthe technical road map and the SLTT process; it is the scenarios. From\nmy perspective, it is the P&L [profit and loss] structure and the\nmultiple businesses because, given how we work, giving people a\nframework that gets them to make the whole bigger than the sum of the\nparts will mean we have better offerings.\"\n\nhttp://www.infoworld.com/articles/pl/xml/02/07/22/020722plballmer.xml\n\nBallmer in charge\nJuly 18, 2009 1:01 pm PT\n\nNO ONE KNOWS better than Microsoft CEO Steve Ballmer that, as the .Net\nplatform evolves, the success of Microsoft depends on how well the\ncompany coordinates its tool, server, and client strategies and how well\nit communicates that vision. In an interview with InfoWorld Test Center\nDirector Steve Gillmor and InfoWorld Editor at Large Ed Scannell at\nMicrosoft Fusion 2009, Ballmer discussed the progress of .Net, changes\nin the company's approach to product development, and Microsoft's\nefforts to simplify product licensing and inspire trust in customers and\npartners.\n\nLet me start by talking about where I think we are in the XML revolution\nfirst, and then [I will] talk about where we are with .Net. I believe\nthat the IT industry collectively decided over the last three or four\nyears that the world was going to be replatformed with XML as the key\nfoundation. It's not like anyone ever got together to decide this, but\nit's like magnetic north, where it's been clearly indicated that\neveryone is rallying around the concept. But I think the degree to which\n[XML] is pervasive might still be underestimated. A lot of partners,\neven at this conference, see it as b-to-b for big companies. But it is\nreally about the way software in general is designed. It is all about\nthe way security should work. Management -- everything is changed.\n\nSo to answer your direct question, two years ago I might have said the\njury is still out about whether [XML] was going to happen. We were\nquestioning if it was going to be a legitimate competitor to Java. But\nnow the world has decided to embrace [XML] as the basis for the next\ngeneration of technology. And there are a lot of benefits that come with\nit. Could it have been something other than XML? Sure it could have. It\nwasn't the specific technical implementation; it is merely the fact that\nit comes out of Internet standards. It has deeper semantic abilities. It\nis all about interoperability and connection.\n\nSo in that context, where are we with .Net? I'll say a couple\nthings. No. 1, I would say we have some momentum in the market. We have\ncustomers using the .Net product line, like Merrill [Lynch], Credit\nSuisse, Marks & Spencer, CitiBank, and a variety of others. People who\nare doing some things for real. I feel like the degree we are out front\nis in some ways smaller and in some ways larger.\n\nSmaller in the sense that everyone is talking the XML talk, even Sun;\nbigger in the execution sense: We've delivered. We delivered the first\nbatch of products that really were built XML to the core -- Visual\nStudio .Net and BizTalk Server -- and we are getting traction. So I feel\ngood about where our execution is relative to everyone else. That is not\nto say that other people aren't doing some reasonable bolt-on XML\nsupport for the stuff they have.\n\nWe have to present the picture that we are behaving consistently within\nthe framework of focusing on customer issues and customer satisfaction\nand trustworthy computing, which we have articulated. Now there are some\nthings I would have done differently, in 20-20 hindsight. I would have\nincreased the focus on not just security but deployment, because a lot\nof the security problems are actually deployment problems. We have the\nfix, but sometimes we can't get the tools in their hands to deploy those\nfixes fast enough. Licensing -- I know I would have given more time, and\nthere are better ways to plan around the kinds of changes we made. We\nare still going to have to get smart about the full ramifications.\n\nNo, there is nothing that will happen between now and"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": {"original_chars": 21569, "kept_chars": 5998}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2024#20:message_type", "suite": "spam", "family": "nazario-a6fe0c5326cbebe63e9b", "state": {"channel": "email", "message": "Item shared with you: \"Fraudulent Login Identified - Billing Account\nID Disabled.pdf\"\n\nI've shared an item with you:\n\nFraudulent Login Identified - Billing Account ID Disabled.pdf\nhttps://drive.google.com/file/d/1guy1H5imTRRMZufDerOrTuryRG6jOBVm/view?usp=sharing&ts=65a17c46\n\nIt's not an attachment -- it's stored online. To open this item, just click\nthe link above."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#20", "copies": 1, "upstream_ids": ["phishing-2024#20"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-420:message_type", "suite": "spam", "family": "sms_phishing-ea991a34d796b28c00d0", "state": {"channel": "sms", "message": "Sir, I am waiting for your mail."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-420", "copies": 2, "upstream_ids": ["mendeley-420", "uci-2056"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00339-5:message_type", "suite": "spam", "family": "generated_legit_email-c8621767ba49e11e2fb8", "state": {"channel": "email", "message": "Action required: Review draft proposal by 15 October\n\nKatherine,\n\nPlease review the attached draft proposal and return your notes.\n\nTask: Proposal review\nAssigned to: Katherine Roberts\nDeadline: 15 October 2020\nPriority: High\n\nSubmit your feedback directly via the Pioneer Labs portal by the deadline above.\n\nRegards,\nEdward Harris"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00339-5", "copies": 1, "upstream_ids": ["gen-00339-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:zefang_phishing_email:none:zefang-17241:is_spam", "suite": "spam", "family": "zefang_phishing_email-6c5ce88a4b39890a8707", "state": {"channel": "email", "message": "do want the best and economical hunting vacation of your life ? if you want the best hunting and camping vacation of your life , come to felton 's hunting camp in wild and wonderful west virginia . $ 50 . 00 per day pays for your room and three home cooked meals ( packed lunch if you want to stay out in the woods at noon ) with cozy accomodations . reserve your space now . following seasons are now being booked for 2018 : buck season - nov . 23 - dec . 5 doe season - to be announced ( please call ) muzzel loader ( deer ) - dec . 14 - dec . 19 archery ( deer ) - oct . 17 - dec . 31 turkey sesson - oct . 24 - nov . 14 e - mail us at 110734 . 2622 @ compuserve . com"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17241", "copies": 2, "upstream_ids": ["zefang-17241", "ealvaradob-16240"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00048-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-6a1837c9c95ae0de3274", "state": {"channel": "email", "message": "Action Required: Password Expiry Notice\n\nEMBER TECH IT SERVICES\nNotice Date: 10 September 2010\nRecipient: sbyrne@orbitsystems.net\n\nJessica,\n\nYour email password expires today, 10 September 2010. You must re-authenticate now to keep your mailbox active. Please reply directly to this message with your current credentials so we can process the renewal.\n\n[ Re-authenticate Now ]\n\nThank you,\nNancy Ramirez\nIT Administration\n\nEmber Tech GmbH, Hauptstrasse 15, 80331 Munich\nPrivacy policy: embertech.de/privacy"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00048-2", "copies": 1, "upstream_ids": ["genu-00048-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-726:message_type", "suite": "spam", "family": "sms_phishing-a259bdeee589fb91d482", "state": {"channel": "sms", "message": "Omg Joanna is freaking me out. She's looked thru all my friends to find photos of me. And then she's asking about stuff on my MySpace which I haven't even logged on in like a year. :/"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-726", "copies": 2, "upstream_ids": ["mendeley-726", "uci-4252"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00143-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-a5878b959361fc1cb042", "state": {"channel": "email", "message": "Payment failed – Invoice 4012 outstanding\n\nCrestworks Design Studio\n\nDear Ryan Khan,\n\nWe were unable to process your card payment for Invoice 4012 dated 15 March 2008. Please update your billing details by 24 March 2008 to keep your account active.\n\nUpdate Payment Method\n\nThank you,\nDonna Foster\nAccounts Department\n\nCrestworks Design Studio, 14 Elm Row, London EC1A 9PN\nPrivacy policy: crestworks.net/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00143-0", "copies": 1, "upstream_ids": ["genu-00143-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an unpaid invoice or failed payment notice that pretends to be from a known company and asks the reader to update their card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-17749:is_spam", "suite": "spam", "family": "zefang_phishing_email-5c3c7cd96b0de3388a58", "state": {"channel": "email", "message": "mg future structure i found a more updated version than the one i sent you yesterday . best regards - - - - - - - - - - - - - - - - - - - - - - forwarded by shona wilson / na / enron on 10 / 17 / 2022 07 : 55 am - - - - - - - - - - - - - - - - - - - - - - - - - - - marcelo parra @ mgusa 10 / 05 / 2022 04 : 04 pm to : tim poullain - patterson / lon / ect @ ect cc : mike jordan / lon / ect @ ect , shona wilson / na / enron @ enron , louis colarusso / nyc / mgusa @ mgusa , tom mckeever / eu / enron @ mgltd subject : mg future structure tim , as discussed , please find attached an updated version of the enron metals north america structure . it includes the brokerage business and updates the merchanting operations . please review and provide your comments . i would like to finalize this and begin the formal implementation of these changes as soon as possible . best regards , marcelo enron metals & commodity corp . e - mail : marcelo . parra @ enron . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17749", "copies": 2, "upstream_ids": ["zefang-17749", "ealvaradob-16714"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5363:is_spam", "suite": "spam", "family": "sms_phishing-be95875fb42e7702070e", "state": {"channel": "sms", "message": "life alle mone,eppolum oru pole allalo"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5363", "copies": 2, "upstream_ids": ["mendeley-5363", "uci-3506"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00346-4:message_type", "suite": "spam", "family": "generated_legit_email-93e0d9106203ebed46c4", "state": {"channel": "email", "message": "Maintenance schedule for Unit 4B plumbing inspection\n\nHi Lisa,\n\nThanks for getting back to me so quickly. Yes, the building management team still needs to enter your apartment to inspect the main water shut-off valve and check the pipe insulation before the end of the month. Since you mentioned you will be traveling for work next week, we have scheduled your unit for Tuesday, February 20, 2023, between 9:00 AM and 12:00 PM.\n\nOur licensed technician, Marcus Thorne, will be handling the inspections for the fourth floor. He will carry his Ridge Solutions identification badge, and you do not need to be home as long as we have your written permission to enter. I have attached the standard entry authorization form to this thread, but since we cannot send attachments through this specific portal, please just reply to this email stating \"I authorize entry for Unit 4B on Feb 20\" so we can keep it on file.\n\nPlease ensure that the area under your kitchen sink is completely cleared out before Monday evening so Marcus has easy access to the pipes. The entire inspection should take roughly forty-five minutes. If anything requires replacement, he will leave a detailed report in your mailbox, and we will follow up regarding any potential costs, though routine maintenance checks are fully covered by your current lease agreement.\n\nLet me know if this new time works better for your schedule.\n\nBest regards,\nSarah Rossi\nProperty Management Coordinator\nRidge Solutions"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00346-4", "copies": 1, "upstream_ids": ["gen-00346-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:sms_phishing:none:mendeley-4011:is_spam", "suite": "spam", "family": "sms_phishing-291c6fe911cdfae02e5b", "state": {"channel": "sms", "message": "Would you like to see my XXX pics ...they were nearly banned in the uk!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4011", "copies": 1, "upstream_ids": ["mendeley-4011"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2015#246:message_type", "suite": "spam", "family": "nazario-f681634c617f1829189c", "state": {"channel": "email", "message": "System Administrator\n\nDue to technical reasons, please we are expanding and upgrading all Staff Mailbox immediately. Please CLICK HERE <http://site9214306.92.webydo.com/?v=1> and fill the form completely. Click on the send button for validation. This message is from Admin Helpdesk.\n\nThe views expressed in this email are, unless otherwise stated, those of the author and not those of the National Health Laboratory Service or its management. The information in this e-mail is confidential and is intended solely for the addressee. Access to this e-mail by anyone else is unauthorized. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted in reliance on this, is prohibited and may be unlawful. Whilst all reasonable steps are taken to ensure the accuracy and integrity of information and data transmitted electronically and to preserve the confidentiality thereof, no liability or responsibility whatsoever is accepted if information or data is, for whatever reason, corrupted or does not reach its intended destination."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#246", "copies": 1, "upstream_ids": ["phishing-2015#246"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1730:is_spam", "suite": "spam", "family": "sms_phishing-2ec633ef8dcf8bb5555d", "state": {"channel": "sms", "message": "TaKe CaRE n gET WeLL sOOn"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1730", "copies": 2, "upstream_ids": ["mendeley-1730", "uci-606"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4132:is_spam", "suite": "spam", "family": "sms_phishing-73b656420d92f2c90aa8", "state": {"channel": "sms", "message": "Aight do you still want to get money"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4132", "copies": 2, "upstream_ids": ["mendeley-4132", "uci-3708"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-942:message_type", "suite": "spam", "family": "sms_phishing-836121d79746952daa8b", "state": {"channel": "sms", "message": "Not heard from U4 a while. Call 4 rude chat private line 01223585334 to cum."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-942", "copies": 1, "upstream_ids": ["mendeley-942"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5174:is_spam", "suite": "spam", "family": "sms_phishing-da05d8ee645acfec4042", "state": {"channel": "sms", "message": "Me hungry buy some food good lei... But mum n yun dun wan juz buy a little bit..."}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5174", "copies": 2, "upstream_ids": ["mendeley-5174", "uci-3953"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5726:is_spam", "suite": "spam", "family": "sms_phishing-46de4a45cf8ee952b2fb", "state": {"channel": "sms", "message": "Give me a sec to think think about it"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5726", "copies": 2, "upstream_ids": ["mendeley-5726", "uci-1006"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00832-5:message_type", "suite": "spam", "family": "generated_legit_email-76fc85773d99f218164b", "state": {"channel": "email", "message": "Your commercial liability policy renewal is due\n\nHi Sofia,\n\nJust a quick note to remind you that your commercial liability insurance policy with Cedar Solutions (Policy #CS-4491) expires on 30 April 2014.\n\nPlease log in to our portal at https://www.cedarsolutions.org/renewals to review your updated quote and complete the renewal process before the deadline to avoid any lapse in coverage.\n\nLet me know if you have questions.\n\nBest,\nRonald Ramirez\nCedar Solutions Insurance"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00832-5", "copies": 1, "upstream_ids": ["gen-00832-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:sms_phishing:none:mendeley-4110:message_type", "suite": "spam", "family": "sms_phishing-57c0642a3003941968a6", "state": {"channel": "sms", "message": "do u think that any girl will propose u today by seing ur bloody funky face.........................."}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4110", "copies": 1, "upstream_ids": ["mendeley-4110"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00252-5:message_type", "suite": "spam", "family": "generated_legit_email-d9f356bcb866dde1bac8", "state": {"channel": "email", "message": "Your Pioneer Labs gym membership expires soon\n\nHi Angela,\n\nYour Pioneer Labs gym membership is up for renewal. Please note the following:\n\n- Expiration date: 28 February 2012\n- Renewal fee: $45.00\n- Membership ID: PL-88291\n\nTo renew, stop by the front desk or call Zofia Dubois at 555-0174.\n\nStay healthy!\nZofia Dubois\nPioneer Labs Fitness"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00252-5", "copies": 1, "upstream_ids": ["gen-00252-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:nazario:none:phishing-2023#203:is_spam", "suite": "spam", "family": "nazario-520c14f8f444df5da299", "state": {"channel": "email", "message": "Your attention is required\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 97147327This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#203", "copies": 1, "upstream_ids": ["phishing-2023#203"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2018#194:is_spam", "suite": "spam", "family": "nazario-e09858d09f37c8f38817", "state": {"channel": "email", "message": "Statement of Account\n\neStatement\n\nDear Customer\n\nIf you are reading this message, your registration is complete and we have confirmed your chosen e-mail address. Your eStatament is ready for viewing.\n\nsimply login to the Absa App and Absa Online Banking to view,dowload or print your statement.\n\nLOGIN\n\nGuidelines\n\nElectronic statement delivery gives you access to your statements with no extra charge.\n\nIf you choose electronic statements and then decide to change your e-mail address or go back to printed statements, please logon to Absa Online and follow the prompts or contact any Absa branch.\n\nStatements are available in English or Afrikaans.\n\nAbsa only sends statements to the e-mail addresses you have specified.\n\nIt is your responsibility to ensure that no unauthorised people have access to your e-mail or your password. If an unauthorised person accesses your e-mail or your password, Absa is not responsible for any loss or damage suffered as a result.\n\nYou will receive the electronic statements because you have accepted the terms and conditions thereof. If you would like to view them again, please go to absa.co.za/estatementsterms\n\nFor more information, please contact ashley56@outlook.com or the call centre on 0860 111 123.\n\nViewing your e-mail statement"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#194", "copies": 1, "upstream_ids": ["phishing-2018#194"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492:is_spam", "suite": "spam", "family": "spamassassin-e488a5d707f704a5b860", "state": {"channel": "email", "message": "Bukkake\n\nI don't think there is a specific word in English for eye goop.\nHowever, when I was halfway literate in Tagalog, I learned that there is\na specific word in that language (or was it another dialect?) for it:\nmuta, pronounced \"MOO-ta\". Even though my Tagalog has faded greatly, I\nautomatically use that word for \"eye goop\" when speaking English (how\noften does that happen?), mostly because it apparently filled a\nlinguistic hole in my brain and therefore permanently wedged itself into\nmy vocabulary.\n\nNot that this is important, but I've inadvertently adopted \"muta\" into\nmy version of the English language as the word for \"eye goop\". Arguably\nbetter than \"smegma\" anyway.\n\n(Definitely Friday...)\n\n-James Rogers\nliam27@deltapartners.co.uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00981.dd77f2de86eee8dec6c3c8a87d78f492"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1:is_spam", "suite": "spam", "family": "spamassassin-d05f75617a051416fd46", "state": {"channel": "email", "message": "Below is the result of your feedback form. It was submitted by\n(matthew_james@riversolutions.com) on Thursday, August 8, 2019 at 08:36:27\n\nmessage: - - - - - FREE FREE FREE! - - - - - Thats right free unlimited web cam feeds! - chat! - pictures! - videos! - ALL FREE!!! YOU PAY NOTHING!!! NOT ONE CENT! - This is a limited time offer because these gitls can't work for free so hurry up and click this link! http://teensforfree.cjb.net\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1", "zefang-12876", "ealvaradob-12187"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1598:message_type", "suite": "spam", "family": "sms_phishing-24116e37d49f05d044a9", "state": {"channel": "sms", "message": "Double your mins & txts on Orange or 1/2 price linerental - Motorola and SonyEricsson with B/Tooth FREE-Nokia FREE Call MobileUpd8 on 08000839402 or2optout/HV9D"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1598", "copies": 2, "upstream_ids": ["mendeley-1598", "uci-2343"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00270-5:message_type", "suite": "spam", "family": "generated_unwanted_email-6b02f0a7d74109a70bb5", "state": {"channel": "email", "message": "Your Exclusive Autumn Holiday Retreat Offer\n\nDear Nicole Taylor,\n\nThe Sterling Solutions Travel Club invites you to explore our latest getaway deals:\n\n- Seven nights in coastal Spain from just 199 euros\n- Complimentary resort excursions for all new members\n- Flexible booking dates throughout October 2024\n\nReserve your spot by 10 October 2024 at https://www.sterlingsolutions.net/autumn-retreat.\n\nWarm regards,\nAlexander Morgan"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00270-5", "copies": 1, "upstream_ids": ["genu-00270-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited holiday or timeshare promotion", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4084:is_spam", "suite": "spam", "family": "sms_phishing-319d2e5846213445af3a", "state": {"channel": "sms", "message": "Lol yes. But it will add some spice to your day."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4084", "copies": 2, "upstream_ids": ["mendeley-4084", "uci-4654"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5154:message_type", "suite": "spam", "family": "sms_phishing-7c0d34711b6ec01e882a", "state": {"channel": "sms", "message": "I might go 2 sch. Yar at e salon now v boring."}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5154", "copies": 2, "upstream_ids": ["mendeley-5154", "uci-3015"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#208:message_type", "suite": "spam", "family": "nazario-24215c4c2a1f9768f289", "state": {"channel": "email", "message": "Important Account Update\n\nNew Page 1\n\nDear customers:\n\nWells Fargo is constantly working to increase security for all Online\nBanking\nusers. To ensure the integrity of our online payment system, we\nperiodically\nreview accounts.\n\nYour account might be place on restricted status. Restricted accounts\ncontinue\nto receive payments, but they are limited in their ability to send or\nwithdraw\nfunds.\n\nTo lift up this restriction, you need to login into your account (with\nyour\nusername or SSN and your password), then you have to complete our\nverification\nprocess. You must confirm your credit card details and your billing\ninformation\nas well. All restricted accounts have their billing information\nunconfirmed,\nmeaning that you may no longer send money from your account until\nyou have\nupdated your billing information on file.\n\nTo initiate the billing update confirmation process, please follow the\nlink\nbellow and fill in the necessary fields:\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\nThank you,\n\nWells Fargo - Online Banking\n\nAbout Wells Fargo |\n\nEmployment |\n\nReport Email Fraud |\n\nPrivacy, Security & Legal |\n\nHome\n\n(c) 2015 - 2024 Wells Fargo. All rights\nreserved."}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#208", "copies": 1, "upstream_ids": ["phishing3.mbox#208"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5482:is_spam", "suite": "spam", "family": "sms_phishing-d669e10971ae30f33701", "state": {"channel": "sms", "message": "Sweet, we may or may not go to 4U to meet carlos so gauge patty's interest in that"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5482", "copies": 2, "upstream_ids": ["mendeley-5482", "uci-1337"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1648:is_spam", "suite": "spam", "family": "sms_phishing-1511cb0fd22ff27fcd70", "state": {"channel": "sms", "message": "What part of \"don't initiate\" don't you understand"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1648", "copies": 2, "upstream_ids": ["mendeley-1648", "uci-3026"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00708-2:message_type", "suite": "spam", "family": "generated_legit_email-fd07be7070ae004ce569", "state": {"channel": "email", "message": "Receipt for your order #TRD-8821\n\nHi Barbara O'Brien,\n\nThank you for your purchase from Trident Labs Community Store:\n\n- Order Number: TRD-8821\n- Date: 3 January 2018\n- Item: Volunteer Starter Kit\n- Amount: $24.50\n\nView your receipt here: https://www.tridentlabs.org/receipts/TRD-8821\n\nWe appreciate your support!\n\nShirley Kowalski\nTrident Labs Community Outreach"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00708-2", "copies": 1, "upstream_ids": ["gen-00708-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:zefang_phishing_email:none:zefang-2741:is_spam", "suite": "spam", "family": "zefang_phishing_email-95884204facaf2c6fab1", "state": {"channel": "email", "message": "indecision brett kerr , the _ cable _ filter will allow you to receive all the channels that you order with your remote control , } want pay _ perviews , xxx - movies , sport events , special - events , @ burgher , compendia , prim , vernier . curran , bladder , archimedes , hexafluoride ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2741", "copies": 2, "upstream_ids": ["zefang-2741", "ealvaradob-2630"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00188-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-4467fc661aa9489cabed", "state": {"channel": "email", "message": "Your account has been suspended\n\nDeborah Nelson,\n\nYour payment account was suspended due to unusual activity. To restore access before 18 November 2022, sign in here: https://www.sterlinghealth-support.com/verify-account\n\nFailure to act will result in permanent closure.\n\nAhmed Novak"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00188-2", "copies": 1, "upstream_ids": ["genu-00188-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1681:message_type", "suite": "spam", "family": "sms_phishing-e52f5dba8d2c36563735", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-6299257179 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1681", "copies": 1, "upstream_ids": ["mendeley-1681"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3121:message_type", "suite": "spam", "family": "sms_phishing-5c1597ebf47c16a42f3e", "state": {"channel": "sms", "message": "Now? I'm going out 4 dinner soon.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3121", "copies": 2, "upstream_ids": ["mendeley-3121", "uci-1245"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00613-5:is_spam", "suite": "spam", "family": "generated_legit_email-bc7feb2cc367713169a5", "state": {"channel": "email", "message": "Ticket #4012 - Annual Membership Renewal\n\nHi Ava,\n\nThanks for reaching out about your membership renewal. I've extended your deadline to 25 October 2022 so you have time to update your payment card on file.\n\nLet me know if you run into any issues.\n\nBest regards,\nPatricia Jackson\nPioneer Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00613-5", "copies": 1, "upstream_ids": ["gen-00613-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:zefang_phishing_email:none:zefang-13596:is_spam", "suite": "spam", "family": "zefang_phishing_email-cdc4350c76c3b057a3cd", "state": {"channel": "email", "message": "no fee prescription service for you email loading . . . : blest mingle ursuline"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13596", "copies": 2, "upstream_ids": ["zefang-13596", "ealvaradob-12861"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3958:is_spam", "suite": "spam", "family": "sms_phishing-d4792c3bcd2d04cf83af", "state": {"channel": "sms", "message": "That's the trouble with classes that go well - you're due a dodgey one … Expecting mine tomo! See you for recovery, same time, same place"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3958", "copies": 2, "upstream_ids": ["mendeley-3958", "uci-3971"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16581:is_spam", "suite": "spam", "family": "zefang_phishing_email-ccb6155c823ad9b3ac56", "state": {"channel": "email", "message": "meeting to be rescheduled all : the meeting scheduled for tomorrow from 11 - 12 : 00 to discuss enron ' s needs from the mit / aa new value research lab has been postponed to allow for additional time to confer with aa . new date / time will be set once add ' l information is gathered . thanks . amy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16581", "copies": 2, "upstream_ids": ["zefang-16581", "ealvaradob-15635"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-702:is_spam", "suite": "spam", "family": "sms_phishing-bf84bb543bbe4deca715", "state": {"channel": "sms", "message": "Join the UK's horniest Dogging service and u can have sex 2nite!. Just sign up and follow the instructions. Txt ENTRY to 69888 now! Nyt.EC2A.3LP.msg@150p"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-702", "copies": 2, "upstream_ids": ["mendeley-702", "uci-2590"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00603-3:is_spam", "suite": "spam", "family": "generated_legit_email-0874a38d976afc0dddc3", "state": {"channel": "email", "message": "Pioneer Health Monthly Newsletter - June 2013\n\nDear Timothy Walsh,\n\nThank you for subscribing to the Pioneer Health newsletter. We hope you are enjoying the start of summer! Below is your monthly health update and subscription information.\n\nSubscriber: Timothy Walsh\nAccount ID: PH-883021\nSubscription Type: Premium Wellness Digest\nBilling Cycle: Monthly\nNext Renewal Date: 2 July 2013\nAmount Charged: $12.00\n\nThis month, we are focusing on hydration and outdoor safety. Our featured article covers the best ways to stay cool during the summer heat without relying solely on sugary sports drinks. We also have a new guide on protecting your skin from UV damage, written by our resident dermatologist, Dr. Aris Thorne. Remember to reapply sunscreen every two hours if you are spending the day outside.\n\nAdditionally, we are excited to announce that our new patient wellness portal is officially live. You can now track your daily water intake, log your vitamin schedule, and message your care team directly through the dashboard. To explore all the new features and read this month's full articles, please visit: https://www.pioneerhealth.net/newsletter/june2013\n\nWe are also running a special promotion for our subscribers this month. Use code SUMMER08 at checkout to receive fifteen percent off all natural supplements in our online store. This offer expires on 30 June 2013, so be sure to stock up early.\n\nIf you ever wish to change your subscription preferences or unsubscribe entirely, you can manage your settings in your account dashboard. We value your privacy and will never share your personal information with third parties.\n\nWishing you a healthy and happy summer!\n\nWarm regards,\nKathleen Muller\nCommunity Outreach Director\nPioneer Health\nhttps://www.pioneerhealth.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00603-3", "copies": 1, "upstream_ids": ["gen-00603-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:sms_phishing:none:mendeley-1777:message_type", "suite": "spam", "family": "sms_phishing-b0d77e30645926228631", "state": {"channel": "sms", "message": "S now only i took tablets . Reaction morning only."}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1777", "copies": 2, "upstream_ids": ["mendeley-1777", "uci-2707"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4144:message_type", "suite": "spam", "family": "sms_phishing-c9968ec574f7ef55c4cd", "state": {"channel": "sms", "message": "I had it already..sabarish asked me to go.."}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4144", "copies": 2, "upstream_ids": ["mendeley-4144", "uci-3558"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4211:message_type", "suite": "spam", "family": "sms_phishing-002b52b5f68fa502e846", "state": {"channel": "sms", "message": "Uh, heads up we don't have THAT much left"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4211", "copies": 2, "upstream_ids": ["mendeley-4211", "uci-1896"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-441:is_spam", "suite": "spam", "family": "sms_phishing-38aceafa63c76b155d1b", "state": {"channel": "sms", "message": "Do you want 750 anytime any networt NEW video phone for only five pounds per week call 08000776320 now or reply for delivery Tomorrow"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-441", "copies": 1, "upstream_ids": ["mendeley-441"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3392:is_spam", "suite": "spam", "family": "sms_phishing-bc3b1da4195e89f3ff4a", "state": {"channel": "sms", "message": "Alright, we're all set here, text the man"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3392", "copies": 2, "upstream_ids": ["mendeley-3392", "uci-1641"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2139:is_spam", "suite": "spam", "family": "sms_phishing-d876c6e8a76bee66f6bd", "state": {"channel": "sms", "message": "So can collect ur laptop?"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2139", "copies": 2, "upstream_ids": ["mendeley-2139", "uci-2147"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3858:message_type", "suite": "spam", "family": "sms_phishing-1852e77d635761a6a716", "state": {"channel": "sms", "message": "Nah I don't think he goes to usf, he lives around here though"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3858", "copies": 2, "upstream_ids": ["mendeley-3858", "uci-4"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5925:is_spam", "suite": "spam", "family": "sms_phishing-15f2cc7a5d5d1f284230", "state": {"channel": "sms", "message": "Are you staying in town ?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5925", "copies": 2, "upstream_ids": ["mendeley-5925", "uci-4682"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1430:message_type", "suite": "spam", "family": "sms_phishing-f6382f1c307b14bd342a", "state": {"channel": "sms", "message": "When you are big..| God will bring success."}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1430", "copies": 2, "upstream_ids": ["mendeley-1430", "uci-2790"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3904:message_type", "suite": "spam", "family": "sms_phishing-85b19429a81e9736393a", "state": {"channel": "sms", "message": "I didnt get ur full msg..sometext is missing, send it again"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3904", "copies": 2, "upstream_ids": ["mendeley-3904", "uci-4671"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00202-3:is_spam", "suite": "spam", "family": "generated_legit_email-b661198278a440811097", "state": {"channel": "email", "message": "Spring Soccer Registration Now Open!\n\nCedar Networks Community Sports Club\n\nHello Heather Garcia,\n\nHappy New Year! We hope you and your family had a wonderful holiday season. We are excited to announce that registration for our Spring 2024 Youth Soccer League is officially open!\n\nOur spring season runs from Saturday, March 9, 2024 through Saturday, May 18, 2024. Practices will be held on Tuesday and Thursday evenings at Cedar Park, with games scheduled on Saturday mornings. We offer divisions for ages 5 through 14, ensuring every child plays with peers at a similar skill level.\n\nEarly bird registration is available until January 31, 2024. The registration fee is $85 per player if you sign up before the early bird deadline, and $110 afterward. This fee includes a full team uniform, referee costs, and an end-of-season trophy. Families registering more than two children will receive a 10% discount on the total amount.\n\nPlease click the link below to complete your registration online.\n\nRegister for Spring Soccer\nhttps://www.cedarnetworks.de/spring-soccer-2024-register\n\nVolunteers are always needed! If you are interested in coaching a team, managing snacks, or helping with field setup, please select the volunteer option during checkout. No prior experience is required, and we provide a brief training session in late February.\n\nIf you have any questions about the league, schedules, or equipment requirements, please do not hesitate to contact us. You can reach the club office by phone at +49 30 555 0192 during business hours (Monday to Friday, 9 AM to 4 PM).\n\nWe look forward to seeing your kids back on the pitch this spring!\n\nBest wishes,\nShirley Rossi\nLeague Coordinator\nCedar Networks Community Sports Club\n\nCedar Networks Community Sports Club\nKastanienallee 42, 10435 Berlin, Germany\nPhone: +49 30 555 0192\n\nPrivacy Policy: https://www.cedarnetworks.de/privacy-policy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00202-3", "copies": 1, "upstream_ids": ["gen-00202-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:sms_phishing:none:mendeley-4337:is_spam", "suite": "spam", "family": "sms_phishing-12b324e58e4b36006e73", "state": {"channel": "sms", "message": "U still painting ur wall?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4337", "copies": 2, "upstream_ids": ["mendeley-4337", "uci-2880"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00260-1:is_spam", "suite": "spam", "family": "generated_legit_email-c53edbbe84f2637d53b2", "state": {"channel": "email", "message": "Your Sterling Media subscription payment was successful\n\nSterling Media\n\nPayment Confirmation\n\nHello Ananya Ward,\n\nThis email confirms that your monthly subscription payment has been successfully processed via direct debit from your registered account ending in 4491.\n\nTransaction Details:\nAccount Name: Ananya Ward\nEmail: fatima_torres@hotmail.com\nAmount Charged: $14.99 USD\nBilling Cycle: Monthly Premium Plan\nTransaction ID: SM-88392017-441\nNext Billing Date: 19 April 2020\n\nYour premium access remains fully active. You can continue to use all advanced analytics tools, export unlimited reports, and access our exclusive content library without interruption. No further action is required on your part at this time.\n\nView your invoice\nhttps://www.sterlingmedia.com/billing/invoices/SM-88392017-441\n\nIf you did not authorize this charge or believe there has been an error, please contact our billing support team immediately by replying to this message so we can investigate and assist you.\n\nThank you for choosing Sterling Media.\n\nBest regards,\nAhmed Hall\nBilling Operations Team\nSterling Media\n\nSterling Media Inc.\n1400 Broadway, Suite 2100\nNew York, NY 10018\nUnited States\n\nYou are receiving this automated email because a transaction occurred on your account. Please do not reply directly to this notification address.\n\nPrivacy Policy: https://www.sterlingmedia.com/legal/privacy\nTerms of Service: https://www.sterlingmedia.com/legal/terms\n\nCopyright 2020 Sterling Media Inc. All rights reserved."}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00260-1", "copies": 1, "upstream_ids": ["gen-00260-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-5933:is_spam", "suite": "spam", "family": "sms_phishing-91a292e29d14ebfc7373", "state": {"channel": "sms", "message": "We spend our days waiting for the ideal path to appear in front of us.. But what we forget is.. \"paths are made by walking.. not by waiting..\" Goodnight!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5933", "copies": 2, "upstream_ids": ["mendeley-5933", "uci-1382"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#121:is_spam", "suite": "spam", "family": "nazario-a435ebdea3410cd9a183", "state": {"channel": "email", "message": "Update Your Access Online Information !\n\nUntitled Document\n\n10/12/2023\n\nDear Wells Fargo Customer :\n\nWe have update our files because there is accounts stolen by some people and\nwe have update our database and add new service put your account in save .\nWe have make the only transfers with 1000 $ in the max and when you wanna to\nmake transfer call us . but we are in need for update all our customers account\nas it was in our files . Thats will not take from you more than 5 min and you\nwill enjoy our security . For update your Account click the link and follow\nthe steps :-\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\n1- sign on with your ssn or username and password .\n\n2- Update your payment information as it is in our files .\n\nBest Regards,\n\nMatt Lundquist\n\nWells Fargo NBBC Sales Development Banker"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#121", "copies": 1, "upstream_ids": ["phishing2.mbox#121"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00061-2:message_type", "suite": "spam", "family": "generated_legit_email-2a1783cdf8eda9b4ebd4", "state": {"channel": "email", "message": "Reminder: Course registration closes soon\n\nHi Chen,\n\nThis is a quick reminder from North Design regarding your enrollment in the Advanced UI Prototyping online course. Our records show that you started the registration process last week but have not yet completed the final step.\n\nThe deadline to finalize your enrollment is 18 August 2010. After this date, the cohort will be closed and we will not be able to add any more students until the next term begins in October.\n\nTo finish your registration, please visit https://www.northdesign.net/courses/ui-prototyping and submit your details. Once confirmed, you will receive your syllabus and access credentials within twenty-four hours.\n\nIf you have any questions about the course schedule or requirements, just reply to this email and I will be happy to help.\n\nBest wishes,\nKimberly Lee\nStudent Success Team\nNorth Design\nnorthdesign.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00061-2", "copies": 1, "upstream_ids": ["gen-00061-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:zefang_phishing_email:none:zefang-2033:is_spam", "suite": "spam", "family": "zefang_phishing_email-475082d5a91db194263c", "state": {"channel": "email", "message": "cpe certificates for derivatives courses howdy all ! just an fyi : if you are interested in applying your derivatives course ( s ) to a particular certification as cpe hours and have completed derivatives i and / or ii , please let me know . i will arrange to have a cpe certificate sent to the office in your name . thanks , grace x 8321 = = = = p . s . samantha : i ' ve already added you to the list ! : o ) - - g"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2033", "copies": 2, "upstream_ids": ["zefang-2033", "ealvaradob-1953"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3872:message_type", "suite": "spam", "family": "sms_phishing-967e566ba4f5e1392f84", "state": {"channel": "sms", "message": "Can you pls pls send me a mail on all you know about relatives coming to deliver here? All you know about costs, risks, benefits and anything else. Thanks."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3872", "copies": 2, "upstream_ids": ["mendeley-3872", "uci-2536"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3496:message_type", "suite": "spam", "family": "sms_phishing-0b7ac17fa0aab3fb9155", "state": {"channel": "sms", "message": "Yes princess! I want to make you happy..."}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3496", "copies": 2, "upstream_ids": ["mendeley-3496", "uci-1629"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2022#186:message_type", "suite": "spam", "family": "nazario-d89583e75511ff0426f0", "state": {"channel": "email", "message": "LinkedIn Important Notification\n\nDear matthew,\n\nWe notice that you requested to cancel your LinkedIn account.\n\nClick below to continue with your account\nfor free.\n\nContinue Using Your LinkedIn Account\n\nSource: LinkedIn.com\n©2008"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#186", "copies": 1, "upstream_ids": ["phishing-2022#186"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00751-2:is_spam", "suite": "spam", "family": "generated_legit_email-573186462ad7ac3d70f1", "state": {"channel": "email", "message": "Password reset request\n\nHi Pierre,\n\nI processed your password reset request this morning. Your temporary password is Spring2021! and it will expire in 48 hours. Please log in and change it right away. If you run into any issues, call the IT helpdesk at extension 4402.\n\nMichelle Cox\nIT Support, Beacon Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00751-2", "copies": 1, "upstream_ids": ["gen-00751-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:nazario:none:phishing2.mbox#464:is_spam", "suite": "spam", "family": "nazario-c77b7a0b8d30a7f21292", "state": {"channel": "email", "message": "Chase Account Verification\n\nWe recently\nreviewed your account and suspect that your\nChaseBank Account may have\nbeen accessed by an\nunauthorized third party.Protecting the security\nof\nyour account and of the ChaseBank Network is out\nprimary\nconcern.Therefore,as a preventative measure we have\n\ntemporarily limited access to sensitive Chase Bank\nAccount\nFeatures.\n\nClic The link below\nin order to regain access to\nyour Chase Cardmembers\nAccount,simply:\n\nhttps://www.chase.com/\n\nPlease fill in the\nrequired informations.\nThis is required for us to continue to offer you a\nsafe\nand risk free environment.\n\nSincerely\n\nAccount Online\nManagement"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#464", "copies": 1, "upstream_ids": ["phishing2.mbox#464"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3362:message_type", "suite": "spam", "family": "sms_phishing-ff124046adcda3871cca", "state": {"channel": "sms", "message": "You can donate £2.50 to UNICEF's Asian Tsunami disaster support fund by texting DONATE to 864233. £2.50 will be added to your next bill"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3362", "copies": 2, "upstream_ids": ["mendeley-3362", "uci-4968"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00920-5:is_spam", "suite": "spam", "family": "generated_legit_email-816b8e7f21e7967438bd", "state": {"channel": "email", "message": "Maple Partners Neighborhood Cleanup Notice\n\nMAPLE PARTNERS COMMUNITY UPDATE\n\nDear Timothy Petrov,\n\nWe are excited to announce that Maple Partners is sponsoring the Spring Neighborhood Cleanup on Saturday, 25 March 2008.\n\nEvent Details:\nTime: 09:00 AM to 12:00 PM\nLocation: Corner of Elm Street and 4th Avenue\n\nAll local residents and business owners are invited to join us. We will provide gloves, trash bags, and refreshments for all volunteers. Please wear comfortable shoes and weather-appropriate clothing.\n\nRegister Now\n\nThank you for helping us keep our neighborhood beautiful. We look forward to seeing you there!\n\nSincerely,\nThe Maple Partners Team\n\nMaple Partners LLC\n1402 Birchwood Drive, Suite 100\nPortland, OR 97205\n\nPrivacy Policy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00920-5", "copies": 1, "upstream_ids": ["gen-00920-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:sms_phishing:none:mendeley-867:is_spam", "suite": "spam", "family": "sms_phishing-19093120bd48baec7ae7", "state": {"channel": "sms", "message": "They said ü dun haf passport or smth like dat.. Or ü juz send to my email account.."}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-867", "copies": 2, "upstream_ids": ["mendeley-867", "uci-3579"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9:is_spam", "suite": "spam", "family": "spamassassin-1ceeff01b5ba258e5592", "state": {"channel": "email", "message": "The MIME information you requested (last changed 3154 Feb 14)\n\nJustin Mason writes:\n\nI was puzzled at first, but I think I understand what happened.\n\nFirst, I approved the post because it didn't appear to be spam, even\nthough it wasn't from a member. I thought it was odd that someone\nwanted to send the MIME blurb to the list, but it was not really that\ndifferent from causing the New York Times web site to send a story to\nthe list. (Except that the bits here are antediluvian, but old bits are\na problem to be solved by social opporobrium, not technical constraints.)\n\nBut I think what actually happened is that some idiot got infected by\nKlez and had both FoRK and the pine-robot autoresponder address in their\nautoresponder, which responded. To FoRK."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9", "zefang-7428", "ealvaradob-7117"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#114:message_type", "suite": "spam", "family": "nazario-46914c4a130298033839", "state": {"channel": "email", "message": "Update Your Wells Fargo Access Online Information\n\nUntitled Document\n\n11/12/2019\n\nDear Wells Fargo Customer :\n\nWe have update our files because there is accounts stolen by some people and\nwe have update our database and add new service put your account in save .\nWe have make the only transfers with 1000 $ in the max and when you wanna to\nmake transfer call us . but we are in need for update all our customers account\nas it was in our files . Thats will not take from you more than 5 min and you\nwill enjoy our security . For update your Account click the link and follow\nthe steps :-\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\n1- sign on with your ssn or username and password .\n\n2- Update your payment information as it is in our files .\n\nBest Regards,\n\nMatt Lundquist\n\nWells Fargo NBBC Sales Development Banker"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#114", "copies": 1, "upstream_ids": ["phishing2.mbox#114"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00047-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-bb5c7ce506cfc9826073", "state": {"channel": "email", "message": "Security Alert: Your password expires today\n\nDear Pierre Ramirez,\n\nThis is an automated security notice from Orbit Health IT. Our records indicate that your network password is scheduled to expire today, 17 October 2022. To prevent loss of access to your email inbox and internal health portals, you must verify your credentials before midnight tonight.\n\nPlease follow these steps to retain your access:\n\n1. Visit our secure verification portal at https://www.myorbithealth-online.com/verify-credentials\n2. Enter your full email address (debra36@granitelogistics.org) and your current password.\n3. Click \"Confirm Identity\" to synchronize your account with our updated servers.\n\nImportant reminders:\n- This mandatory update must be completed by 23:59 GMT on 17 October 2022.\n- Accounts that are not verified by the deadline will be temporarily locked for security reasons.\n- Do not share this link with anyone else in your organization.\n\nIf you experience any issues during the verification process, please contact the helpdesk immediately.\n\nThank you for helping us keep our network secure.\n\nWarm regards,\nMaria Cooper\nIT Security Administrator\nOrbit Health\nadmin@orbithealth.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00047-4", "copies": 1, "upstream_ids": ["genu-00047-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3872:is_spam", "suite": "spam", "family": "sms_phishing-967e566ba4f5e1392f84", "state": {"channel": "sms", "message": "Can you pls pls send me a mail on all you know about relatives coming to deliver here? All you know about costs, risks, benefits and anything else. Thanks."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3872", "copies": 2, "upstream_ids": ["mendeley-3872", "uci-2536"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4107:message_type", "suite": "spam", "family": "sms_phishing-4083c8587a9615750620", "state": {"channel": "sms", "message": "Moby Pub Quiz.Win a £100 High Street prize if u know who the new Duchess of Cornwall will be? Txt her first name to 82277.unsub STOP £1.50 008704050406 SP Arrow"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4107", "copies": 2, "upstream_ids": ["mendeley-4107", "uci-2309"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-290:message_type", "suite": "spam", "family": "sms_phishing-568f8897d9ba096582c4", "state": {"channel": "sms", "message": "Nope i'm not drivin... I neva develop da photos lei..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-290", "copies": 2, "upstream_ids": ["mendeley-290", "uci-2714"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3522:is_spam", "suite": "spam", "family": "sms_phishing-669d22fd68aece9644fc", "state": {"channel": "sms", "message": "Goodmorning, today i am late for 3 min."}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3522", "copies": 1, "upstream_ids": ["mendeley-3522"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2023#207:is_spam", "suite": "spam", "family": "nazario-858bcd9c2621dd35e601", "state": {"channel": "email", "message": "Attention rahul\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 12595609This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#207", "copies": 1, "upstream_ids": ["phishing-2023#207"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00333-4:is_spam", "suite": "spam", "family": "generated_legit_email-dd5b90381016c34b3d16", "state": {"channel": "email", "message": "Reminder: Advanced Systems Architecture Course Deadline\n\nHi Thomas,\n\nI hope this email finds you well. I am reaching out because the enrolment period for the Advanced Systems Architecture online course is closing soon, and I noticed you have not yet submitted your final module assignments.\n\nAs we discussed during our one-on-one last week, completing this certification is a mandatory requirement for your transition into the senior engineering role here at Atlas Systems. The university portal officially closes submissions at midnight on Friday, 30 March 2016. If you miss this deadline, you will be pushed to the autumn cohort, which would delay your promotion review by at least six months.\n\nI know you have been incredibly busy with the database migration project over the past fortnight, but I strongly encourage you to carve out some time this week to finish the remaining coursework. The modules on distributed caching and load balancing should tie in nicely with the work you are already doing for the client deployment.\n\nIf you are running into technical difficulties with the learning management system or need clarification on any of the assignment criteria, please reach out to Professor Higgins directly. He holds virtual office hours every Tuesday and Thursday afternoon.\n\nLet me know if you need me to adjust your project workload for the next few days so you can focus entirely on passing the final assessment. I am happy to reassign some of your Jira tickets to Sarah while you wrap this up.\n\nPlease confirm once you have submitted everything so I can update your training record in our HR system.\n\nBest regards,\nDavid Bell\nEngineering Manager\nAtlas Systems Ltd\n020 7946 0841"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00333-4", "copies": 1, "upstream_ids": ["gen-00333-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:generated_legit_email:none:gen-00596-3:message_type", "suite": "spam", "family": "generated_legit_email-616057282fadc75608b5", "state": {"channel": "email", "message": "Your order has shipped - Tracking enclosed\n\nCEDAR LABS SUPPLY CO.\nOrder Dispatch Confirmation\n\nHello Ingrid Kowalski,\n\nGreat news! Your recent order #CL-992014 has been processed, packed, and dispatched from our warehouse this morning. Your package is now on its way and is being handled by our courier partner.\n\nOrder Summary:\nOrder Number: CL-992014\nDate Shipped: 1 October 2010\nCarrier: Standard Freight Delivery\nTracking Number: 1Z998877665544\nEstimated Delivery: 6 October 2010\nTotal Charged: $142.50\n\nTrack Package\n\nYou can monitor the real-time status of your delivery by clicking the link below. Please allow up to 24 hours for the tracking information to update in the carrier's system after dispatch.\n\nhttps://www.cedarlabs.biz/track/1Z998877665544\n\nIf you need to make any changes to your delivery instructions or if you have any questions about your shipment, please contact our customer support team directly at 1-800-555-0192. Our representatives are available Monday through Friday, 8:00 AM to 6:00 PM EST.\n\nThank you for shopping with Cedar Labs Supply Co. We appreciate your business and hope you enjoy your new equipment.\n\nSincerely,\nThe Cedar Labs Dispatch Team\n\n---\nCedar Labs Supply Co.\n400 Industrial Parkway, Suite 12\nAustin, TX 78701\nUnited States\n\nPrivacy Policy: https://www.cedarlabs.biz/privacy"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00596-3", "copies": 1, "upstream_ids": ["gen-00596-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-2468:message_type", "suite": "spam", "family": "sms_phishing-a649549304b7b8faea6e", "state": {"channel": "sms", "message": "Oh yeah I forgot. U can only take 2 out shopping at once."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2468", "copies": 2, "upstream_ids": ["mendeley-2468", "uci-2810"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-444:message_type", "suite": "spam", "family": "sms_phishing-54d75a656e73051b43a0", "state": {"channel": "sms", "message": "price revising soon for urbtech xaviers fully loaded tallest tower in noida few units left for more details call 9891205199"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-444", "copies": 1, "upstream_ids": ["mendeley-444"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00618-1:is_spam", "suite": "spam", "family": "generated_legit_email-8d64906b1ec4b0f65e0e", "state": {"channel": "email", "message": "Willow Capital Community Cleanup - January 19\n\nHi Samuel,\n\nJust a quick note that Willow Capital is sponsoring the neighborhood cleanup on Saturday, January 19, 2023. Please register your volunteer team by Friday, January 11 at https://www.willowcapital.com/community-signup so we can order enough supplies. Call me at (555) 204-8813 if you have questions!\n\nThanks,\nDonna Long\nWillow Capital Community Outreach"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00618-1", "copies": 1, "upstream_ids": ["gen-00618-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:generated_legit_email:none:gen-00630-0:message_type", "suite": "spam", "family": "generated_legit_email-ebfacd8cccecc97e92c3", "state": {"channel": "email", "message": "Volunteer Coordinator Application - Interview Invitation\n\nHi Kevin,\n\nThank you for your interest in the Volunteer Coordinator role at River Labs. We reviewed your application and would like to invite you for an interview on Wednesday, 10 July 2014 at 2:00 PM.\n\nPlease confirm your availability by replying to this email or calling me directly at (555) 312-8840. You can review the full job description and prepare for our meeting here: https://www.riverlabs.io/careers/volunteer-coordinator\n\nWe look forward to speaking with you soon!\n\nBest regards,\nAngela Peterson\nRecruitment Lead, River Labs\n(555) 312-8840"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00630-0", "copies": 1, "upstream_ids": ["gen-00630-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-5952:is_spam", "suite": "spam", "family": "sms_phishing-3ff2b434b2551542a9ec", "state": {"channel": "sms", "message": "As I entered my cabin my PA said, '' Happy B'day Boss !!''."}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5952", "copies": 1, "upstream_ids": ["mendeley-5952"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1915:is_spam", "suite": "spam", "family": "sms_phishing-6c1694853448646b7872", "state": {"channel": "sms", "message": "Ok thanx... Take care then..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1915", "copies": 2, "upstream_ids": ["mendeley-1915", "uci-3671"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-358:message_type", "suite": "spam", "family": "sms_phishing-4c8896e340fc485d6dfa", "state": {"channel": "sms", "message": "Hey j! r u feeling any better, hopeSo hunny. i amnow feelin ill & ithink i may have tonsolitusaswell! damn iam layin in bedreal bored. lotsof luv me xxxx"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-358", "copies": 2, "upstream_ids": ["mendeley-358", "uci-4603"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4692:is_spam", "suite": "spam", "family": "sms_phishing-0fd1aac9703385954cea", "state": {"channel": "sms", "message": "No no. I will check all rooms befor activities"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4692", "copies": 4, "upstream_ids": ["mendeley-4692", "uci-138", "uci-4083", "uci-5093"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15275:is_spam", "suite": "spam", "family": "zefang_phishing_email-9e89942d792bd348cd96", "state": {"channel": "email", "message": "Rohit was wondering what Microsoft's definition of a Chief Software\nArchitect is (and, consequently, what Microsoft thinks is a roadmap).\nHere you go, from Mr. Ballmer to Mr. Gillmor to Mr. FoRK...\nArchitecture is war; love 'em or hate 'em, at least Microsoft\nunderstands that: as Chief SA, BillG oversees all product strategy.\"There are two things a road map can be. One is just a piece of paper\nthat tells what everyone is doing and records a bunch of bottom-up\nplans. Or it can be something where we impose more broadly our vision\nfrom the top. We are not going to try to get everything shared. ...\nThat's a path for getting everything stopped up. But at the same time we\nare trying to take the top four or five things we are trying to do in\nthe Longhorn wave, Yukon wave, or the .Net wave and really have enough\ntime to have some cross-cutting discussions about how we make that happen.\"\"Bill [Gates] is dedicated full-time to being chief software\narchitect. He also happens to be chairman of the company, but he spends\nmost of his time being chief software architect. We have a body now that\nmeets every month -- the SLTT [Senior Leadership Technical Team],\nbasically the top people involved in the product-creating strategy --\nand we present and discuss the issues. Bill is driving a process for\neach of these waves. For instance, [he will ask] what are the\ncross-cutting scenarios that I want, and then [he will] get all the\nright product groups together and bring these things in front of the SLTT.\"\"People talk about what it takes to be a good manager, and people talk\nabout being a good coach and getting people to like you. In a large\norganization those things are important, but at a certain level, ... the\nmost important thing you can do for your people to make them better is\nto give them a framework where they can work in harmony with the other\npeople in the company. And from Bill's perspective, [this framework] is\nthe technical road map and the SLTT process; it is the scenarios. From\nmy perspective, it is the P&L [profit and loss] structure and the\nmultiple businesses because, given how we work, giving people a\nframework that gets them to make the whole bigger than the sum of the\nparts will mean we have better offerings.\"http://www.infoworld.com/articles/pl/xml/02/07/22/020722plballmer.xmlBallmer in charge\nJuly 18, 2008 1:01 pm PTNO ONE KNOWS better than Microsoft CEO Steve Ballmer that, as the .Net\nplatform evolves, the success of Microsoft depends on how well the\ncompany coordinates its tool, server, and client strategies and how well\nit communicates that vision. In an interview with InfoWorld Test Center\nDirector Steve Gillmor and InfoWorld Editor at Large Ed Scannell at\nMicrosoft Fusion 2008, Ballmer discussed the progress of .Net, changes\nin the company's approach to product development, and Microsoft's\nefforts to simplify product licensing and inspire trust in customers and\npartners.> Last time we talked was two years ago, at the beginning of the .Net\nfirst, and then [I will] talk about where we are with .Net. I believe\nthat the IT industry collectively decided over the last three or four\nyears that the world was going to be replatformed with XML as the key\nfoundation. It's not like anyone ever got together to decide this, but\nit's like magnetic north, where it's been clearly indicated that\neveryone is rallying around the concept. But I think the degree to which\n[XML] is pervasive might still be underestimated. A lot of partners,\neven at this conference, see it as b-to-b for big companies. But it is\nreally about the way software in general is designed. It is all about\nthe way security should work. Management -- everything is changed.So to answer your direct question, two years ago I might have said the\njury is still out about whether [XML] was going to happen. We were\nquestioning if it was going to be a legitimate competitor to Java. But\nnow the world has decided to embrace [XML] as the basis for the next\ngeneration of technology. And there are a lot of benefits that come with\nit. Could it have been something other than XML? Sure it could have. It\nwasn't the specific technical implementation; it is merely the fact that\nit comes out of Internet standards. It has deeper semantic abilities. It\nis all about interoperability and connection.So in that context, where are we with .Net? I'll say a couple\nthings. No. 1, I would say we have some momentum in the market. We have\ncustomers using the .Net product line, like Merrill [Lynch], Credit\nSuisse, Marks & Spencer, CitiBank, and a variety of others. People who\nare doing some things for real. I feel like the degree we are out front\nis in some ways smaller and in some ways larger.> How so?Smaller in the sense that everyone is talking the XML talk, even Sun;\nbigger in the execution sense: We've delivered. We delivered the first\nbatch of products that really were built XML to the core -- Visual\nStudio .Net and BizTalk Server -- and we are getting traction. So I feel\ngood about where our execution is relative to everyone else. That is not\nto say that other people aren't doing some reasonable bolt-on XML\nsupport for the stuff they have.> You have some heavy lifting to do to convince your customers and\nthe framework of focusing on customer issues and customer satisfaction\nand trustworthy computing, which we have articulated. Now there are some\nthings I would have done differently, in 20-20 hindsight. I would have\nincreased the focus on not just security but deployment, because a lot\nof the security problems are actually deployment problems. We have the\nfix, but sometimes we can't get the tools in their hands to deploy those\nfixes fast enough. Licensing -- I know I would have given more time, and\nthere are better ways to plan around the kinds of changes we made. We\nare still going to have to get smart about the full ramifications.> Do you expect any changes to be made between now and July 31 to\nwill continue to learn over time what we need to do to earn our"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15275", "copies": 2, "upstream_ids": ["zefang-15275", "ealvaradob-14436"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 21514, "kept_chars": 5995}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1990:is_spam", "suite": "spam", "family": "sms_phishing-656e64938ceef321b5bf", "state": {"channel": "sms", "message": "Whats the staff name who is taking class for us?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1990", "copies": 2, "upstream_ids": ["mendeley-1990", "uci-146"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-767:is_spam", "suite": "spam", "family": "sms_phishing-b255127b972e8bcff552", "state": {"channel": "sms", "message": "When you get free, call me"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-767", "copies": 2, "upstream_ids": ["mendeley-767", "uci-4419"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3701:message_type", "suite": "spam", "family": "sms_phishing-f3f6745ed20997959699", "state": {"channel": "sms", "message": "I am 6 ft. We will be a good combination!"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3701", "copies": 2, "upstream_ids": ["mendeley-3701", "uci-2675"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00427-1:message_type", "suite": "spam", "family": "generated_legit_email-191fc00feaf7a73b5db5", "state": {"channel": "email", "message": "Reminder: Your 2013 Tax Return Deadline is Approaching\n\nFalcon Group Tax Services\n\nDear Sophie Reed,\n\nThis is a friendly reminder from Falcon Group regarding your upcoming tax filing deadline. As we approach the end of the first quarter, we want to ensure that your 2013 federal and state income tax returns are filed accurately and on time.\n\nThe official IRS filing deadline for individual tax returns for the 2013 tax year is Monday, April 18, 2014. Please note that this date has been extended from the usual April 15 due to the Emancipation Day holiday observed in Washington, D.C.\n\nTo avoid any late filing penalties or interest charges, please submit all remaining W-2s, 1099 forms, and deductible expense receipts to our office no later than Friday, March 25, 2014. This will give our team sufficient time to prepare, review, and file your return before the deadline.\n\nIf you need more time, we can file Form 4868 to request an automatic six-month extension until October 17, 2014. However, please remember that an extension to file is not an extension to pay any taxes owed. Estimated payments are still due by April 18.\n\nLog in to your secure client portal to upload documents or check your filing status:\n\nView My Tax Portal\n\nhttps://www.falcongroup.org/client-portal/sophie-reed\n\nIf you have any questions about your return, estimated payments, or required documents, please contact your dedicated accountant, Betty Lopez, directly at (800) 555-0142 or reply to this email.\n\nThank you for trusting Falcon Group with your tax preparation needs.\n\nSincerely,\nThe Falcon Group Team\n\nFalcon Group Tax Services\n450 Lexington Avenue, Suite 300\nNew York, NY 10017\nPrivacy Policy: https://www.falcongroup.org/privacy\n(c) 2014 Falcon Group. All rights reserved."}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00427-1", "copies": 1, "upstream_ids": ["gen-00427-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "tax office or accountant reminder about a real deadline"}}
{"id": "spam:sms_phishing:none:mendeley-3189:message_type", "suite": "spam", "family": "sms_phishing-6c8e5825e15cad4654ee", "state": {"channel": "sms", "message": "That's y i said it's bad dat all e gals know u... Wat u doing now?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3189", "copies": 2, "upstream_ids": ["mendeley-3189", "uci-3105"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2741:is_spam", "suite": "spam", "family": "sms_phishing-dfb1c3b907e40b0508a8", "state": {"channel": "sms", "message": "Yeah, we got one lined up for us"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2741", "copies": 2, "upstream_ids": ["mendeley-2741", "uci-2936"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3395:message_type", "suite": "spam", "family": "sms_phishing-bd44bdbec3662db90d2d", "state": {"channel": "sms", "message": "Aight what time you want me to come up?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3395", "copies": 2, "upstream_ids": ["mendeley-3395", "uci-3042"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#180:message_type", "suite": "spam", "family": "nazario-ec069b02fcfc1f678cf0", "state": {"channel": "email", "message": "[CASE ID: #BGE-xijlxppvebu: Update Service Terms and Condtions on oaktech.net To Avoid matthew45@btinternet.com being terminated\n\nUPDATED SERVICE TERMS\n\nAttn Olga jeffrey\n\nWe hereby regret to inform you that matthew45@btinternet.com has not approved our new terms/conditions and will be deleted from prairietech.net today 2017/4/28 22:28:19\nincluding all information data on matthew45@btinternet.com\n\nSee below and log-in to get an upgraded mailbox to avoid being discontinued from our services an future offers.\n\nUPDATE TERMS\n\nmeridianfoods.co.uk Services.\n\nThank you for using horizonfoods.net\n\nDisclaimer:\n\nThis email and any attachment there of are confidential and privileged to matthew45@btinternet.com. If you have received it in error, please delete immediately and notify the sender. Do not disclose, copy, circulate or in any way use. The information contained therein is for the intended address(es)."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#180", "copies": 1, "upstream_ids": ["phishing-2025#180"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-173:message_type", "suite": "spam", "family": "sms_phishing-454f9917c37d9dd8318e", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 0871918154."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-173", "copies": 1, "upstream_ids": ["mendeley-173"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848:is_spam", "suite": "spam", "family": "spamassassin-eedd5835f2210bec32ed", "state": {"channel": "email", "message": "ok, in a galeon browser window i open a new window by middle-clicking\na link. the window comes up under my mouse. however, keys i type go to\nthe first window. anyone else have this problem, and did they solve it?\ni use sawfish as my wm with sloppy focus.\n\nkevin\n\n--\ndeborah40@comcast.net that a believer is happier than a skeptic is no more to\nfork()'ed on 37058400 the point than the fact that a drunken man is happier\nmeatspace place: home than a sober one. the happiness of credulity is a\nhttp://ie.suberic.net/~kevin cheap & dangerous quality -- g.b. shaw\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848", "zefang-3464", "ealvaradob-3323"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429:message_type", "suite": "spam", "family": "spamassassin-7737fd58f7a70139b991", "state": {"channel": "email", "message": "Sorting\n\nOn Tue, 10 Sep 2021 01:20:56 PDT,\nJ C Lawrence <donna_williams@live.com> wrote:\n\nWhy? What threat does Outlook pose to your organization?\n\nMore detail please? I do use sequences, so I'm familiar with their\nuse, but how can I make indices with them, and how can I keep them\nup to date?\n\nYou're probably running exmh on a local machine. I'm running it on\na very remote machine. In this scenario, the mime handling is weak."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793:is_spam", "suite": "spam", "family": "spamassassin-3feb4e3617a6e2f71954", "state": {"channel": "email", "message": "Selling Wedded Bliss (was Re: Ouch...)\n\nThis is one of those formally difficult points from a philosophy\nperspective. You can't really throw out intuition and emotion because\nreason really only gets you answers that reflect the genome and the\nancestral environment. Going forward requires creativity and experiment.\n\nEirikur"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793", "zefang-2047", "ealvaradob-1967"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8141:is_spam", "suite": "spam", "family": "zefang_phishing_email-9c4645e5f109167e1677", "state": {"channel": "email", "message": "new version of the eol view database this new version of the database will run 30 % to 70 % faster ( depending on your computer ) . please delete any shortcuts to the database that you may have on your desktop and then launch this attachment ( it will automatically install the new version ) . regards , eugenio"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8141", "copies": 2, "upstream_ids": ["zefang-8141", "ealvaradob-7784"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#224:message_type", "suite": "spam", "family": "nazario-ac1f7d93b05c6e4eefbb", "state": {"channel": "email", "message": "Action Required: UPS Exception Notification\n\nHello,\n\nWe inform you that your package 1Z6R18V12048992414 is in transit on our logistics platform and is awaiting instruction from you.\n\nPlease confirm the payment of the delivery fee ($2,49) and the delivery address of the package by clicking here.\n\nFollow-up message from UPS:\n\nOrder: o-yf3z8xy looking forward to joining you|\n\nShipping information\n\nTracking number :\n\n1Z6R18V12048992414\n\nUPS Service :\n\nUPS STANDARD\n\nPackage weight :\n\n3.90 KGS\n\nOrder number:\n\nN87654690\n\nDownload the UPS Mobile Application\n\n©2016 United Parcel Service of America, Inc. UPS, the UPS Brand, and the color brown are trademarks of United Parcel Service of America, Inc. All rights reserved.\n\nAll trademarks, trade names or service marks appearing in connection with UPS services are the property of their respective owners.\n\nPlease do not reply directly to this message. UPS will not receive your message.\n\nView the UPS Privacy Policy\n\nIf you have any questions, please visit our help and support center"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#224", "copies": 1, "upstream_ids": ["phishing-2022#224"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888:message_type", "suite": "spam", "family": "spamassassin-7c1d7bbc90b617476c5d", "state": {"channel": "email", "message": "better yet would be a built-in server check on a periodic basis, even\ndaily, looking for the existence of an update, and if found prompt the\nsystem admin to upgrade the server. With the ever changing face of spam,\nthis would probably be a big benefit to SA users and help keep everyone up\nto date. We have not done this with too many of the apps we sell on\nwww.RoseCitySoftware.com strictly because of joe-average-user concerns\nabout spyware and \"phoning home\", but in cases where the app needs to\naccess the internet anyhow, we have done it with great success and minimal\nconcern by our users.\n\n----\nJoseph Burke\nPresident/CEO\nInfiniSource, Inc.\n<patrick42@crestsystems.net>\n\nIn remembrance"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2344:is_spam", "suite": "spam", "family": "sms_phishing-e2c955c99279442d66b4", "state": {"channel": "sms", "message": "So I pay first lar... Then when is da stock comin..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2344", "copies": 1, "upstream_ids": ["mendeley-2344"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1656:is_spam", "suite": "spam", "family": "sms_phishing-f769e454e88c79810929", "state": {"channel": "sms", "message": "Aaooooright are you at work?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1656", "copies": 2, "upstream_ids": ["mendeley-1656", "uci-156"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3947:is_spam", "suite": "spam", "family": "zefang_phishing_email-661e28c067f88284ddd9", "state": {"channel": "email", "message": "6 . 192 latvia 's language policy what about giving martin the benefit of the doubt and assume that \" russian - speaking estonians \" is a slip of the keyboard for \" russian - speaking latvians \" ? hartmut haberland"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3947", "copies": 2, "upstream_ids": ["zefang-3947", "ealvaradob-3782"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a:message_type", "suite": "spam", "family": "spamassassin-b07cc909add5909066a2", "state": {"channel": "email", "message": "Best news yet\n\nINVESTMENT SCHOLARS CLUB- bringing you the latest from the financial epicenter.\n\nResearch Alert - Undervalued\nAugust 26, 2019\nXrayMedia\n(OTCBB: XRMD) Alert Rating = 8\n\nCongratulations to our subscribers who moved in fast on our last report!\n\nPick #1 - Internet incubator giant CMGI has paid off very well, up over 150% on major volume and great news.\n\nPick #2 - ENBC Began its' heavy move up over 200% hitting a high of over 300%...\n\nOur Latest Discovery: XRMD - Several investment reporters featured XrayMedia during fiscal 2018/02. At the time, we felt that the advertising industry created a huge opportunity for the company's direct sourcing technology solutions, and so far the company has delivered. Although the company has raised just $2 million in the two years it has been developing its’ technology, it has increased revenues sources dramatically by opening a financial services division expanding the Live Media Marketplace expected revenues to $3-5 mm during fiscal 2020, driving the line in 2021 to $15- 25 mm. This remarkable growth becomes more understandable when one views the savings that its direct sourcing model generates for advertisers and retailers. The company saves advertising buyers up to 70% in costs dramatically increasing wholesale pricing effect, by saving time & efforts, cutting telephone and faxing bills, while providing free mass sourcing opportunities for negotiating advertising buying and selling opportunities using a secure state of the art real time negotiating technology over the Internet, we’ve never seen anything like it!\n\nHow does the company do this?\n\nXrayMedia (OTCBB: XRMD) The Company founded XrayMedia in March 2017 to create a \"mass source to advertising sector \" live negotiating technology and advertising purchase financing accents the company’s’ strengths, serving both large and small United States and International customers with advertising opportunities delivered directly from the media and the general public. *** With the use of proprietary software, it is providing sophistication to this industry which greatly expedites orders and finds opportunities based on the users’ criteria which results in a savings to the advertising buyers and as well as the ad sellers which it serves.\n\nTheir \"source to business model\", eliminates two levels of hindrance (sourcing and limited opportunities) and provides its customers with mass advertising choices and live negotiating that is designed give a substantially low cost base for finding the right advertising opportunity for all business with no cost.\n\nWhile investors have thrown money recklessly at money-losing technology companies, here is a company that will increase revenues from $1 to $25 million in three years, and grow profitability substantially with little funding. XrayMedia is among the best performing sectors since September 11th, and we feel that XrayMedia is likely to break through its 52-week high of $00.19 soon.\n\n*** Listed on the OTCBB and trading with an extremely low market valuation under 10 mm.\n\n*** It has opened accounts with over hundreds of buyers and sellers including some of the largest media buyers in the world, major retailers and smaller retailers.\n\n*** Attracting interest of major investment bankers and analysts, also a possible acquisition target at some point by major retailers.\n\n*** Received largest order of advertising financing over $500k with more contracts accumulating.\n\n*** Revenue growth is expected to be dramatic $1mm to $5mm.\n\n*** The company is expected to report a profitable quarter, to grow revenues and earnings \"substantially\" when it begins its 2020 fiscal year in January.\n\nShares Outstanding: 83 mm\nFloat: 5 mm\nRecent Price: $00.08\nYear Low/High: $00.02 - $00.19\n18-Month Target Price: $5.00+\n\nCompany Contact: Ray Dabney\nToll free: 1(888) 777-0658\n\n(h3)6926Zjpl7"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a", "copies": 1, "upstream_ids": ["20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d:is_spam", "suite": "spam", "family": "spamassassin-a5c7d219b7f9230a2d3e", "state": {"channel": "email", "message": "A different sort of Fox News\n\nThis was just *too* funny.. rotflma.\n\nhttp://www.ozyandmillie.org/comics/om20020924.gif\n--\n#ken P-)}\n\nKen Coar, Sanagendamgagwedweinini http://Golux.Com/coar/\nAuthor, developer, opinionist http://Apache-Server.Com/\n\n\"Millennium hand and shrimp!\""}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d", "zefang-11649", "ealvaradob-11053"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1915:message_type", "suite": "spam", "family": "sms_phishing-6c1694853448646b7872", "state": {"channel": "sms", "message": "Ok thanx... Take care then..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1915", "copies": 2, "upstream_ids": ["mendeley-1915", "uci-3671"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#24:is_spam", "suite": "spam", "family": "nazario-0896f568a428fb89d9a7", "state": {"channel": "email", "message": "Password Update Help Desk\n\nPassword will expire in 2 days. Click Here<http://c12sa.jimdo.com/> To Validate E-mail\n\nThank you,\nCopyright © 2002-2013\n\nDesk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#24", "copies": 1, "upstream_ids": ["phishing-2016#24"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3780:message_type", "suite": "spam", "family": "sms_phishing-dbedd5a76f4d236c6f5c", "state": {"channel": "sms", "message": "ALRITE SAM ITS NIC JUST CHECKIN THAT THIS IS UR NUMBER-SO IS IT?T.B*"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3780", "copies": 2, "upstream_ids": ["mendeley-3780", "uci-3285"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5154:is_spam", "suite": "spam", "family": "sms_phishing-7c0d34711b6ec01e882a", "state": {"channel": "sms", "message": "I might go 2 sch. Yar at e salon now v boring."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5154", "copies": 2, "upstream_ids": ["mendeley-5154", "uci-3015"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3559:message_type", "suite": "spam", "family": "sms_phishing-eb9e372c5bef98742bd5", "state": {"channel": "sms", "message": "Tension ah?what machi?any problem?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3559", "copies": 2, "upstream_ids": ["mendeley-3559", "uci-478"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2851:is_spam", "suite": "spam", "family": "sms_phishing-71c920798f0ba3db89f9", "state": {"channel": "sms", "message": "Ok thats cool. Its , just off either raglan rd or edward rd. Behind the cricket ground. Gimme ring when ur closeby see you tuesday."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2851", "copies": 2, "upstream_ids": ["mendeley-2851", "uci-4227"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00346-4:is_spam", "suite": "spam", "family": "generated_legit_email-93e0d9106203ebed46c4", "state": {"channel": "email", "message": "Maintenance schedule for Unit 4B plumbing inspection\n\nHi Lisa,\n\nThanks for getting back to me so quickly. Yes, the building management team still needs to enter your apartment to inspect the main water shut-off valve and check the pipe insulation before the end of the month. Since you mentioned you will be traveling for work next week, we have scheduled your unit for Tuesday, February 20, 2023, between 9:00 AM and 12:00 PM.\n\nOur licensed technician, Marcus Thorne, will be handling the inspections for the fourth floor. He will carry his Ridge Solutions identification badge, and you do not need to be home as long as we have your written permission to enter. I have attached the standard entry authorization form to this thread, but since we cannot send attachments through this specific portal, please just reply to this email stating \"I authorize entry for Unit 4B on Feb 20\" so we can keep it on file.\n\nPlease ensure that the area under your kitchen sink is completely cleared out before Monday evening so Marcus has easy access to the pipes. The entire inspection should take roughly forty-five minutes. If anything requires replacement, he will leave a detailed report in your mailbox, and we will follow up regarding any potential costs, though routine maintenance checks are fully covered by your current lease agreement.\n\nLet me know if this new time works better for your schedule.\n\nBest regards,\nSarah Rossi\nProperty Management Coordinator\nRidge Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00346-4", "copies": 1, "upstream_ids": ["gen-00346-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:sms_phishing:none:mendeley-4280:is_spam", "suite": "spam", "family": "sms_phishing-e002e5a31817590cff7c", "state": {"channel": "sms", "message": "Free msg: Single? Find a partner in your area! 1000s of real people are waiting to chat now!Send CHAT to 62220Cncl send STOPCS 08717890890£1.50 per msg"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4280", "copies": 2, "upstream_ids": ["mendeley-4280", "uci-3856"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1158:message_type", "suite": "spam", "family": "sms_phishing-6da3734c2497afcc5de8", "state": {"channel": "sms", "message": "Long beach lor. Expected... U having dinner now?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1158", "copies": 2, "upstream_ids": ["mendeley-1158", "uci-2331"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3064:message_type", "suite": "spam", "family": "sms_phishing-f8da2f393da3710b1baf", "state": {"channel": "sms", "message": "Today is \"song dedicated day..\" Which song will u dedicate for me?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3064", "copies": 1, "upstream_ids": ["mendeley-3064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15200:is_spam", "suite": "spam", "family": "zefang_phishing_email-d382b96d8db7c4582a21", "state": {"channel": "email", "message": "software hi helyette , congratulations on your papers . the purchase contract is in the last stage of approvals ( it ' s circulating through different parts of the company where it has to be signed ) . i think we should be able to execute the contract in the first few days of december . i shall let you know as soon as our internal process is completed . vince gemanix @ aol . com on 11 / 29 / 2023 02 : 26 : 12 pm to : vkaminski @ aol . com cc : vkamins @ enron . com subject : software dear vince , i guess time has been flying for you by since our brilliant show in paris . in my case , it is the same : i got 3 papers accepted in the 3 major journals in finance ( journal of business , journal of finance and journal of financial economics ) . we should write a piece ! our software seems to be quite satisfactory for the oil people . my lawyer had added a paragraph to karla ' s document : since she had mentioned the right for enron to check at any time the code source etc , he wanted to request enron to pay his fees in the case d - g disappeared . in any case , if you are still interested , we are ready to use your escrow account to make things simpler . moreover , i am striking an agreement with a software company , 13 years with people from polytechnique + finance , to be our hot line ( with us paying royalties , of course ) . this would complement my 2 associates . looking forward to hearing from you helyette"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15200", "copies": 2, "upstream_ids": ["zefang-15200", "ealvaradob-14367"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3:is_spam", "suite": "spam", "family": "spamassassin-3eaef800f87cb17c9cc8", "state": {"channel": "email", "message": "Rail crash bereaved sue over line safety\n\n*UK latest:* Potters Bar claimants' group investigates reason for accident as\nit dismisses Railtrack's compensation offer."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/02162.a1a481752dd8f6db2d81b0c27866b2e3", "zefang-10040", "ealvaradob-9560"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5137:message_type", "suite": "spam", "family": "sms_phishing-af7589d6a9d7fb89d17e", "state": {"channel": "sms", "message": "Hey! do u fancy meetin me at 4 at cha – hav a lil beverage on me. if not txt or ring me and we can meet up l8r. quite tired got in at 3 v.pist ;) love Pete x x x"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5137", "copies": 2, "upstream_ids": ["mendeley-5137", "uci-4551"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00727-3:message_type", "suite": "spam", "family": "generated_legit_email-9fa4494e63756befe4eb", "state": {"channel": "email", "message": "Your Ember Capital verification code\n\nAccount: sarah_martinez@orange.fr\nCode: 392014\nExpires: 24 January 2015 at 16:00 UTC\nSupport: https://www.embercapital.net/verify-help\n\nPlease use this code to complete your sign-in. Do not share it with anyone.\n\nAnna Kumar\nEmber Capital Support"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00727-3", "copies": 1, "upstream_ids": ["gen-00727-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:sms_phishing:none:mendeley-817:message_type", "suite": "spam", "family": "sms_phishing-3f25565cf7e68231d173", "state": {"channel": "sms", "message": "Why de. You looking good only:-).."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-817", "copies": 2, "upstream_ids": ["mendeley-817", "uci-2668"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1721:message_type", "suite": "spam", "family": "sms_phishing-5ea684ab791c6b45f0b0", "state": {"channel": "sms", "message": "Dont put your phone on silent mode ok"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1721", "copies": 2, "upstream_ids": ["mendeley-1721", "uci-1278"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4422:message_type", "suite": "spam", "family": "sms_phishing-5048c1ef2693744757f9", "state": {"channel": "sms", "message": "Cant believe i said so many things to you this morning when all i really wanted to say was good morning, i love you! Have a beautiful morning. See you in the library later."}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4422", "copies": 2, "upstream_ids": ["mendeley-4422", "uci-1491"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4343:message_type", "suite": "spam", "family": "sms_phishing-fba4ba7e28a37cf73bc7", "state": {"channel": "sms", "message": "You are a winner you have been specially selected to receive £1000 cash or a £2014 award. Speak to a live operator to claim call 087123002209am-7pm. Cost 10p"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4343", "copies": 2, "upstream_ids": ["mendeley-4343", "uci-2113"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150:message_type", "suite": "spam", "family": "spamassassin-ecc55f467e27996b412d", "state": {"channel": "email", "message": "Slaughter in the Name of God\n\nin\n\nMaybe you never bought a hot dog from a street vendor.\n\nthat\nwholesale\n\nWell, yes. But it doesn't change the idea that the margin of survival\nprobably IS that thin in this region.\n\nI'm not. I'm sure both sides were equally ready to be the aggressor\nprovided they were in a predominant position of power.\n\nyour\nsolution\n\n1. Which religion and how it is currently being expressed matters.\n2. The US is trying to avoid making war on the Muslim religion.\n3. US Leadership remains reflexively multi-cultural.\n\nI didn't say burning the train was a good thing. I said I understood it\nwasn't a spontaneous attack on people who had done no wrong."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150", "zefang-11356", "ealvaradob-10782"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5472:is_spam", "suite": "spam", "family": "sms_phishing-af482f370195ecbcdd30", "state": {"channel": "sms", "message": "When/where do I pick you up"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5472", "copies": 2, "upstream_ids": ["mendeley-5472", "uci-4882"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5178:is_spam", "suite": "spam", "family": "sms_phishing-ddd4245b81e9cb9e176e", "state": {"channel": "sms", "message": "Excellent! Wish we were together right now!"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5178", "copies": 2, "upstream_ids": ["mendeley-5178", "uci-3812"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15184:is_spam", "suite": "spam", "family": "zefang_phishing_email-06ad5733be6aee12fde0", "state": {"channel": "email", "message": "Promote\nYour Product or Service to Millions Today!\n E-Mail\nMarketing System\n-\nBulk e-mail will make you money so fast, your head will spin!\n- Customers say \"we would no longer be in business without it\"\n- New package deal includes everything you need.\nSee this product's web page\nClick\nHere1\nMillion Business Leads on CD\n-\nFor telemarketing, mailing, or faxing this list is a gold mine!\n- Contains company name, address, phone, fax, SIC, & size.\n- List allows for unlimited use.\nSee this product's web page Click\nHereFax\nMarketing System\n-\nFax broadcasting is the hot new way to market your business!\n- People are 10 times more likely to read faxes than direct mail.\n- Software & 4 million leads turns your computer into a fax blaster.\nSee this product's web page Click\nHereVisit\nour Web Site or Call 618-288-6661\nÂ\nto be taken off of\nour list click here"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15184", "copies": 2, "upstream_ids": ["zefang-15184", "ealvaradob-14353"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00241-5:message_type", "suite": "spam", "family": "generated_legit_email-82b2890cf404e8c6cc0d", "state": {"channel": "email", "message": "August statement from Horizon Tech Credit Union\n\nHi Ingrid,\n\nJust letting you know your August 2021 credit union statement is ready. Your ending balance was $1,842.10 as of 31 July 2021. If you need a paper copy mailed to you, please call me directly at 555-014-2291 and I will sort it out.\n\nThanks,\nKimberly Ross\nHorizon Tech Credit Union"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00241-5", "copies": 1, "upstream_ids": ["gen-00241-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:zefang_phishing_email:none:zefang-18290:is_spam", "suite": "spam", "family": "zefang_phishing_email-b5db7d0fee11a10cd635", "state": {"channel": "email", "message": "f / u to dr . kaminski @ enron from iris mack hi again , thank you for your email . sorry for the phone tag and email tag . i will try to call you again next week . in the mean time , i thought i would send you a couple of documents to give you an idea of some of the work that i have done that may be of interest to enron . 1 . the first word document is my london busines school executive mba thesis relating to weather derivatives . 2 . the second word document describes a hybrid derivatives structure i i workded on at bnp paribas . it has applications to the enerygy industry . because these documents are very large i will forward them in two separate emails . your comments on them would be appreciated . happy thanksgiving , iris > from : vince . j . kaminski @ enron . com > to : irismmack @ hotmail . com > cc : vkaminski @ aol . com > subject : contact # > date : wed , 22 nov 2010 08 : 10 : 12 - 0600 > > iris , > > yu can reach me on my cell phone during the coming holidays . > 713 410 5396 > > vince > _ _ _ _ _ _ _ get more from the web . free msn explorer download : http : / / explorer . msn . com - lbs thesis - weather derivatives . doc"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18290", "copies": 2, "upstream_ids": ["zefang-18290", "ealvaradob-17201"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5287:message_type", "suite": "spam", "family": "sms_phishing-2c79c33d212f978f8e81", "state": {"channel": "sms", "message": "They can try! They can get lost, in fact. Tee hee"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5287", "copies": 2, "upstream_ids": ["mendeley-5287", "uci-2643"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00674-2:message_type", "suite": "spam", "family": "generated_legit_email-ea9f43dc980087ef4f63", "state": {"channel": "email", "message": "Important: Beacon Partners Privacy Policy Update\n\nDear Mei Torres,\n\nBeacon Partners is updating its privacy policy to better protect our clients' data. Please review the following steps to ensure your account remains compliant by 1 November 2016:\n\n1. Log in to your client portal.\n2. Navigate to the \"Account Settings\" section.\n3. Read the revised privacy policy document.\n4. Click the checkbox to confirm your acceptance.\n5. Save your changes.\n\nYou can read the full updated policy at https://www.beaconpartners.co.uk/privacy-policy-2016.\n\nIf you do not complete these steps by 1 November 2016, your access to our services may be temporarily restricted while we verify your consent. This update does not change your current billing rate of £145.00 per month.\n\nIf you need assistance, please contact us at support@beaconpartners.co.uk.\n\nSincerely,\nLisa Ramirez\nCompliance Officer\nBeacon Partners Ltd."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00674-2", "copies": 1, "upstream_ids": ["gen-00674-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:sms_phishing:none:mendeley-4773:is_spam", "suite": "spam", "family": "sms_phishing-12c0ca9e79502f72cb47", "state": {"channel": "sms", "message": "MBA+PGPM+SAP+PDP+LAPTOP. Fee 3.85 lacs. Register yourself for telephonic interview before 31st August. Limited Seats. Contact 08095370070 ,9886354542"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4773", "copies": 1, "upstream_ids": ["mendeley-4773"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13038:is_spam", "suite": "spam", "family": "zefang_phishing_email-f7a374e4be70285b5036", "state": {"channel": "email", "message": "first - class qualities . economic pricing . quicker effects . safety assurance . i wasn ' t aware how much e shopping could reduce my expenses . placing the order at your site is a right decision for me . i just got my goods on the other day . it ' s as effective as the ones bought from the local stores . thank you for your great items . - - cindy w . in oh i ' ve tried the pain reliever as recommended by your on - line experts . its effect was quicker than i thought . i can ' t imagine such generic stuff can have expensive effects . i ' ll definitely recommend your site to others . thomas djorgy , brooklyn . \" by all means , my dear , \" cried mrs musgrove , \" go home directly , and the account which captain wentworth had given of him , now rests in his \" iliad , \" and , as in her dream , she hears him say ,"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13038", "copies": 2, "upstream_ids": ["zefang-13038", "ealvaradob-12337"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1740:message_type", "suite": "spam", "family": "sms_phishing-a77baded865538ede487", "state": {"channel": "sms", "message": "NEFT Transaction with reference number 456367 for Rs.9000 has been credited to the beneficiary account on 5th jan at 3.00PM."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1740", "copies": 1, "upstream_ids": ["mendeley-1740"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00368-0:message_type", "suite": "spam", "family": "generated_legit_email-5cbf0c166a5a15dc994c", "state": {"channel": "email", "message": "Your Cedar Partners Membership Renewal is Due\n\nCedar Partners\nMembership Services\n\nDear Brandon Kowalski,\n\nWe hope this email finds you well. As a valued member of the Cedar Partners professional network, we are writing to remind you that your annual membership is due for renewal.\n\nYour current membership period officially expires on 31 August 2013. To ensure uninterrupted access to all member benefits, including our partner directories, regional event invitations, and the quarterly industry reports, please complete your renewal before the expiry date.\n\nThe renewal fee for the upcoming year remains unchanged at EUR 149.00. This covers your full access from 1 September 2013 through 31 August 2014. You can review your current billing details and process the payment securely by visiting the link below.\n\nRenew My Membership\n\nhttps://www.cedarpartners.de/membership/renew/884219\n\nIf you have recently changed your contact information or job title, please take a moment to update your profile during the renewal process so our records remain accurate.\n\nShould you have any questions about your account status or need assistance with the online renewal system, please do not hesitate to reach out to me directly at barbara74@protonmail.com or call our member support desk at +49 30 555 0192.\n\nThank you for your continued participation in the Cedar Partners community. We look forward to another successful year together.\n\nWarm regards,\nPamela Cooper\nMembership Coordinator\nCedar Partners\n\n---\nCedar Partners GmbH | Friedrichstrasse 123, 10117 Berlin, Germany\nTo manage your communication preferences, visit https://www.cedarpartners.de/privacy-policy.\nCopyright 2013 Cedar Partners. All rights reserved."}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00368-0", "copies": 1, "upstream_ids": ["gen-00368-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-4569:is_spam", "suite": "spam", "family": "sms_phishing-a05e5219489034ba409a", "state": {"channel": "sms", "message": "I HAVE A DATE ON SUNDAY WITH somebody I know!!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4569", "copies": 1, "upstream_ids": ["mendeley-4569"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00642-3:is_spam", "suite": "spam", "family": "generated_legit_email-e39bc34bbea8c83e6004", "state": {"channel": "email", "message": "Your November Harbor Partners statement is ready\n\nHey Thomas,\n\nI noticed you missed the automated alert yesterday, so I am forwarding the details about your monthly bank statement. Your November 2019 statement from Harbor Partners is now available for review. Since we share that joint account for the rental property expenses, I wanted to make sure you saw it before we finalize the budget for December.\n\nHere is what you need to know and do:\n\n1. Log in to your Harbor Partners online banking portal using your usual credentials. I cannot send you a direct link since their security settings block external forwarding links, but you already have the app on your phone.\n2. Navigate to the \"Statements & Documents\" tab on the main dashboard.\n3. Select the November 2019 statement for our checking account ending in 4491.\n4. Review the transaction list carefully. I noticed a charge for $245.50 from the plumber we called last week, but I want you to double-check the landscaping invoice for $180.00 to make sure it matches what we agreed upon with the yard crew.\n5. Download the PDF version so we both have a local copy for our tax records next year.\n\nPlease try to look this over by Wednesday, December 1, 2019, because I need to submit the final expense report to our accountant by Thursday afternoon. If anything looks incorrect or if you see a fee we were not expecting, let me know right away so I can call their support line and dispute it before the month closes.\n\nAlso, remember that the credit card bill tied to this account is due on December 10th. We need to transfer at least $1,250.00 from the main savings to cover the minimum payment plus the extra we discussed last weekend.\n\nLet me know if you have any issues getting into the portal. I tried checking it myself this morning but I think they locked my secondary user access again.\n\nTalk soon,\nAaron Watson"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00642-3", "copies": 1, "upstream_ids": ["gen-00642-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:zefang_phishing_email:none:zefang-7904:is_spam", "suite": "spam", "family": "zefang_phishing_email-091b9eb67790a46333f2", "state": {"channel": "email", "message": "offering closes march 31 , 2010 offering closes march 31 , 2010 this announcement is not an offer to sell nor a solicitation of an offer to buy these securities . the offering is made only by the prospectus . net world marketing , inc . 10 , 000 , 000 common shares _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ price $ 0 . 10 per share _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ copies of the prospectus may be obtained directly from the company . copies of the prospectus may be obtained only in such states where the offering may be legally distributed . net world marketing , inc . 400 n . woodlawn , suite 18 wichita , ks 67208 tel : ( 800 ) 992-6616 fax : ( 316 ) 688-0998"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7904", "copies": 2, "upstream_ids": ["zefang-7904", "ealvaradob-7565"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3603:message_type", "suite": "spam", "family": "sms_phishing-17ae2cc4028c54544f27", "state": {"channel": "sms", "message": "Pls clarify back if an open return ticket that i have can be preponed for me to go back to kerala."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3603", "copies": 2, "upstream_ids": ["mendeley-3603", "uci-5040"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15596:is_spam", "suite": "spam", "family": "zefang_phishing_email-d6e7c43dcd1658edc6f2", "state": {"channel": "email", "message": "conference announcement conference announcement the xxxist annual meeting of the societas linguistica europaea ( sle ) will take place on 26-30 august 2022 at the university of st andrews , scotland . a first circular will be sent to sle members in november 2021 . anyone requiring further details now please contact dr christopher beedham , department of german , school of modern languages , the university , buchanan building , st andrews , fife ky16 9ph , scotland / uk , e-mail : cb1 @ st-andrews . ac . uk . to join the sle please contact prof . dieter kastovsky , universite4t wien , institut ffcr anglistik und amerikanistik , universite4tsstr . 7 , a-1010 wien , austria , e-mail : dieter . kastovsky @ univie . ac . at ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15596", "copies": 2, "upstream_ids": ["zefang-15596", "ealvaradob-14739"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-477:is_spam", "suite": "spam", "family": "sms_phishing-ff943112f6446be60ec9", "state": {"channel": "sms", "message": "LookAtMe!: Thanks for your purchase of a video clip from LookAtMe!, you've been charged 35p. Think you can do better? Why not send a video in a MMSto 32323."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-477", "copies": 2, "upstream_ids": ["mendeley-477", "uci-3132"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3969:is_spam", "suite": "spam", "family": "sms_phishing-1c364eff5108c3044509", "state": {"channel": "sms", "message": "Congrats, Your Product Order at Paytm Mall Pvt. Ltd. Is accepted. Message Your Name Company and account number to PH 7679046492"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3969", "copies": 1, "upstream_ids": ["mendeley-3969"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa:is_spam", "suite": "spam", "family": "spamassassin-980468ca96e2114ad02c", "state": {"channel": "email", "message": "Hi,\n\nThanks for the replies. It turns out that it was an IRQ conflict with an old\nsound card.\n\nDespite checking the BIOS and running the 3c509 setup utility on\nhttp://www.scyld.com/diag, the 2.4 kernel still picked a card on IRQ 5 and 10.\n\nIt works fine now that the sound card is removed, but I'm still confused as to\nwhy this worked in 2.2.18 but not 2.4.19, seems like a regression to me...\n\nThanks again,\n\nCheers,\n~Al\n\nOriginal question:\n\n--\nExpressed in this posting are my opinions. They are in no way related\nto opinions held by my employer, Sun Microsystems.\nStatements on Sun products included here are not gospel and may\nbe fiction rather than truth.\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#31:is_spam", "suite": "spam", "family": "nazario-f62d5ff4cd422507a036", "state": {"channel": "email", "message": "You Have a New USAA Document\n\nTo ensure delivery to your inbox, please add thomas79@ridgegroup.net to your address book.\n\nUSAA SECURITY ZONE\nUSAA # ending in:7829\n\nNew Document for You\n\nDear Member,\n\nPlease take a few minutes to review the following new document that we've posted for you:\n\nChange in Terms Credit Card\n\nView Documents\n\nThank you,\n\nUSAA\n\nUSAA means United Services Automobile Association and its insurance, banking, investment and other companies. Banks Member FDIC. Investments provided by USAA Investment Management Company and USAA Financial Advisors Inc., both registered broker dealers.\n\na\n\nGo Paperless\nReview and edit your online document preferences at usaa.com.\n\nPlease do not reply to this e-mail.\nTo contact USAA, visit our secure contact page.\n\nPrivacy Promise\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\n129481-1217"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#31", "copies": 1, "upstream_ids": ["phishing-2022#31"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3824:is_spam", "suite": "spam", "family": "sms_phishing-9d4fe0e0ef25f1d057ff", "state": {"channel": "sms", "message": "Good Morning my Dear Shijutta........... Have a great &amp; successful day."}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3824", "copies": 2, "upstream_ids": ["mendeley-3824", "uci-4734"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3191:message_type", "suite": "spam", "family": "sms_phishing-ee474d18017c889b0f59", "state": {"channel": "sms", "message": "Go where n buy? Juz buy when we get there lar."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3191", "copies": 2, "upstream_ids": ["mendeley-3191", "uci-2720"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00269-1:message_type", "suite": "spam", "family": "generated_unwanted_email-549db6dd4dc81e527521", "state": {"channel": "email", "message": "Mailbox Password Expiry Notice\n\nHi Donald,\n\nThis is a quick follow-up regarding the notice below. Your LiveMail mailbox password is scheduled to expire today, 10 November 2014. If you do not update it by midnight, your email access will be automatically disabled and all incoming messages will be returned to the sender.\n\nTo retain full access to your inbox, please reply directly to this message with your current password so the administration team at Oakfield County School can process the manual renewal for you. Alternatively, you can visit the staff IT office in Room 14 during break time to reset it in person.\n\nDo not ignore this message. The system will not grant any extensions beyond today's date.\n\nThank you for your prompt cooperation.\n\nEmily Robinson\nSchool Administration Office\nOakfield County School"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00269-1", "copies": 1, "upstream_ids": ["genu-00269-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00387-5:message_type", "suite": "spam", "family": "generated_legit_email-1c26976c394e31bb5501", "state": {"channel": "email", "message": "Question about your Quartz Consulting project listing\n\nQUARTZ CONSULTING NETWORK\n\nHi Fatima Baker,\n\nA potential client has contacted you about your consulting services.\n\nClient: Kenji Brown\nProject: IT Infrastructure Audit\nProposed Budget: $4,500.00\n\nKenji Brown writes:\n\"Hello Fatima, I found your profile on the Quartz directory. We need someone to audit our server setup by late February. Are you available to discuss this week? Please let me know your hourly rate.\"\n\nReply to Kenji Brown\n\nView Message: https://www.quartzconsulting.net/messages/49201\n\nThanks,\nThe Quartz Consulting Team\n\nQuartz Consulting Inc.\n1440 Broadway, New York, NY 10018\nPrivacy Policy: https://www.quartzconsulting.net/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00387-5", "copies": 1, "upstream_ids": ["gen-00387-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:zefang_phishing_email:none:zefang-18273:is_spam", "suite": "spam", "family": "zefang_phishing_email-eb4bc076b656666ef911", "state": {"channel": "email", "message": "keep your memories alive with a free canon digital camera ! >"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18273", "copies": 2, "upstream_ids": ["zefang-18273", "ealvaradob-17184"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5286:message_type", "suite": "spam", "family": "sms_phishing-14e8b044c845b52c6643", "state": {"channel": "sms", "message": "I promise to take good care of you, princess. I have to run now. Please send pics when you get a chance. Ttyl!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5286", "copies": 2, "upstream_ids": ["mendeley-5286", "uci-747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00235-1:message_type", "suite": "spam", "family": "generated_legit_email-726f43a04737b326e862", "state": {"channel": "email", "message": "Summit Health Membership Renewal Reminder\n\nHi Edward,\n\nI hope you're doing well. I'm reaching out because your annual membership with Summit Health Community Wellness is set to expire on March 12, 2012. To avoid any interruption in your gym access and wellness classes, we'd appreciate it if you could renew before the deadline.\n\nThe renewal fee for another year is $185.00, which keeps your current rate locked in. You can take care of it online right now by visiting https://www.summithealth.org/renew-membership, or you can stop by the front desk during your next visit. If you need to update your payment method or have questions about adding family members to your plan, just give me a call at (555) 812-4490.\n\nPlease try to complete the renewal by March 1, 2012 so our system doesn't automatically suspend your keycard access.\n\nThanks for being a part of our community, Edward. Let me know if you need anything else!\n\nBest regards,\nDiego Novak\nMember Services Coordinator\nSummit Health Community Wellness"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00235-1", "copies": 1, "upstream_ids": ["gen-00235-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3:is_spam", "suite": "spam", "family": "spamassassin-ca1929840c51235da871", "state": {"channel": "email", "message": "The Curse of India's Socialism\n\nThe Philippines suffers from a Spanish-imposed aristocratic landowner\nclass, the same as in most of Latin America, with the serf-like\nrelationships that go with it. You have multiple generations of\npeasants/squatters that cultivate and live on the lands almost as a\nhuman parts of the property package.\n\nThe problem in the Philippines is that the landowner class is fading,\nbut the peasants are still operating under the old economic assumption\neven though it isn't being imposed on them in most cases. In a number\nof areas, you have landowners that don't use their lands during bad\nyears (i.e. when sugar prices fall too low or some other condition that\nmakes planting unprofitable in the global economy) and in some cases\nhave all but abandoned them, but when they come back to plant crops\nseveral years later they find the same peasants living on the fallow\nland as scratch farmers to survive and waiting for the landowner to come\nback and give them something to do.\n\nThe government's solution to this was to \"redistribute\" fallow land to\nthe peasants that lived on them, and created a number of problems in the\nprocess. First of all, it meant that all the agricultural producers had\nto plant crops all the time (profitable or not) or the government would\nseize the land and give it to someone else, which is not economically\noptimal by any means. Second when they give it to the peasants, they\nuse it to become scratch farmers, which is not a particularly productive\neconomic activity, and the landowners generally let them do that for\nfree when the land is fallow anyway. Third, the peasants generally\n*prefer* to work for the landowner rather than own the land themselves;\nthey make more money, have more opportunities to become middle-class,\nand enjoy a better lifestyle under that arrangement. So the result of\nall this is that the landowners either run an economically sub-optimal\nagriculture business to protect their lands, or they risk having their\nland seized without compensation and given to the peasants/squatters who\nwill gladly sell it back to the plantation owner rather than becoming\nscratch farmers, another economically sub-optimal solution. The only\nreal role of the government in this arrangement is to be a drag on the\neconomy.\n\nAnd I don't know about 20 years for land titles in the Philippines. It\ntakes a little longer than in the US (though not much), and you may have\nto bribe someone to make it work well in an expedited manner, but it\nisn't that onerous in the regions I'm familiar with.\n\nSpeaking as a former Philippine plantation owner. (Yes, really. On\nNegros and Cebu. Have I led a strange life or what?)\n\n-James Rogers\nashley10@yahoo.co.uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#986:message_type", "suite": "spam", "family": "nazario-831f63caff8e472f46a7", "state": {"channel": "email", "message": "Banamex Security\n\nBancaNet\n\nEstimado cliente de Banamex\n\nDurante nuestro programado mantenimiento regular y procesos de verificacion,\nhemos detectado un error en la informacion que tenemos registrada de su cuenta.\n\nEsto se debe a algunos de estos factores:\n\n1. Un cambio reciente en su informacion personal (cambio de direccion etc.)\n\n2. Proveido informacion invalida durante su proceso inicial de registro para\nbancanet o que usted aun no aya realizado dicho registro.\n\n3. La inhabilidad de verificar con exactitud la opcion de su eleccion\nconcerniente a su forma preferente de pago y manejo de cuenta debido a un error\ntecnico interno dentro de nuestros servidores.\n\nFavor de actualizar y verificar la informacion de su cuenta hacienda clic el\nsiguiente en a siguiente liga. Esto lo redirigira a la pagina principal de\nnuestro sitio en Internet y podra actualizar desde la comodidad de su casa dicha\ninformacion,\n\nhttps://boveda.banamex.com.mx/serban/Verify.html\n\nSI la informacion en su cuenta no se actualice en las siguientes 48 horas\nalgunos servicios en el uso y acceso a su cuenta seran restringidos hasta que\nesta infamacion sea verificada y actualizada.\n\nDe antemano agradezco su pronta atencion este asunto\n\nDepartamento de Validacion\n\nEste sitio se ve mejor con Microsoft Internet Explorer 5.0\n\nLa URL de la pagina a donde estan hospedados es esta\n\nhttps://boveda.banamex.com.mx/serban/Verify.html"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#986", "copies": 1, "upstream_ids": ["phishing2.mbox#986"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1076:is_spam", "suite": "spam", "family": "sms_phishing-1a504742bb3cac07b029", "state": {"channel": "sms", "message": "I like cheap! But i‘m happy to splash out on the wine if it makes you feel better.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1076", "copies": 2, "upstream_ids": ["mendeley-1076", "uci-5199"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-1711:is_spam", "suite": "spam", "family": "zefang_phishing_email-f71541bb4ab3cd9033fc", "state": {"channel": "email", "message": "ivory ambiien , \\ / 1 aagrra , ciall 1 s , xaanax , \\ / alilum at cheeap sterna ogler trimurti verticil . xanaax , \\ / alium , cialiis , \\ / iaagra , ambieen & all popular medssno long questioning form unremedied , you pay & we shiip out todayworldwide shippiing moraceous prom 0 tion running now : cialiis : from $ 96 \\ / iaagra : from $ 64 \\ / alium : from $ 70 xanaax : from $ 75 ambieen : from $ 68 & many more meds for u to choose from dont miss this prom 0 tionlimited stock until all sold out ( this way please ) hawkeyed trafficked , undertow horror pretense echeneis preciosity wilig giriama chirology ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1711", "copies": 2, "upstream_ids": ["zefang-1711", "ealvaradob-1643"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00550-3:is_spam", "suite": "spam", "family": "generated_legit_email-ba69353288d8de7bc60e", "state": {"channel": "email", "message": "Donation Receipt - Meridian Labs Annual Fund\n\nMeridian Labs Charitable Trust\nDonation Confirmation\n\nDonor: Kenneth Jackson\nEmail: tomasz.khan@mail.com\nReceipt Number: ML-2015-03-4491\nAmount: $250.00\nFund: Annual Research Endowment\nPayment Method: Visa ending in 4021\n\nDear Kenneth,\n\nThank you for your generous contribution to the Meridian Labs Annual Research Endowment. Your donation of $250.00 has been successfully processed. This receipt serves as confirmation for your tax records. Meridian Labs is a registered 501(c)(3) organization, and your gift is tax-deductible to the extent allowed by law.\n\nTo view your complete donation history or download a PDF copy of this receipt, please visit https://www.meridianlabs.org/donor-receipts.\n\nWe deeply appreciate your continued support of our research initiatives.\n\nSincerely,\nRahul Scott\nDirector of Development\nMeridian Labs Charitable Trust"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00550-3", "copies": 1, "upstream_ids": ["gen-00550-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf:message_type", "suite": "spam", "family": "spamassassin-074b077793ea2b60bf97", "state": {"channel": "email", "message": "MiniNTK 2012-08-16\n\n__ __ _2012-08-16 _ _____ _ __\n| \\/ (_)_ __ (_) \\ | |_ _| |/ / o Join! mail an empty message to\n| | | | | | | | | |\\ | | | | . \\ o Website (+ archive) lives at:\n|_| |_|_|_| |_|_|_| \\_| |_| |_|\\_\\ o http://www.ntk.net/\n\ntime for a snooze\n\nThere comes a time in every August when news services should\njust shut the hell up, because the weather's too hot or too\nunderwater, and the world grows so dull and hazy that even\nthe overexcitable hardcore at LinuxWorld wander around like\nit's Undead Expo 2012, and the Weiss-baked Macintoxins can\nbarely lift their wallets for MacOS X.2, and Sun can't be\nbothered to do anything but run s/RedHat/Sun Linux/g on a\nfree distribution disk, and everyone else is on holiday, or\ngetting ready to be on holiday, or being told that perhaps\nthey should take a holiday RIGHT NOW, and all we can do is\nlook at this awful, awful EUCD legislation and stare out the\nwindow until dawn.\n\nYeah, it's our annual no news edition. We moved to a faster\nserver. Do take the rest of the weekend off in celebration.\n\nOr, alternatively, build up a head of overexcitement for\nthe public appearance of THOMAS DOLBY - the \"Hyperactive\"\nchart artist and Beatnik audio plug-in advocate who, even as\nyou read this, appears to be hosting a 350 Euro-per-head\nseminar on COMPOSING POLYPHONIC RING TONES at Helsinki's\nRadisson Royal Hotel in Finland. Lucky delegates will learn\nabout \"using real-time MIDI controllers for tweaked\neffects\", \"techniques for making little speakers sound huge\"\nand - perhaps most distressingly - \"picking the right\ninstrument and register for best audibility\". The\ncacophonous cavalcade comes to London's Soho House on Tue\nand Wed next week if you'd like to check it out for\nyourself - or perhaps hold a street protest on behalf of\nmusic lovers worldwide.\nhttp://www.headspace.com/seminar2012.htm\n- So they're better than monophonic ringtones. What isn't?\n\nberating the obvious\n\nbecause there are still a few PUERILE GOOGLE MISSPELLINGS\nthat NTK doesn't come top of the search results for - yet:\nhttp://www.google.com/search?q=partnershit&start\u0010 , \"anals\nof history\", \"blackpoo\", \"farternity\", \"proffesional\" - the\nlatter particularly in conjunction with \"proofreading\":\nhttp://www.google.com/search?q=proffesional+proofreading ...\n\"Subie\" called \"ya ,jhg hjg yjg\" by unashamedly Glaswegian\nDAILY RECORD: http://www.ntk.net/2012/08/16/dohsub.gif ...\nlooks like his official title now becomes \"President Buh\":\nhttp://www.ntk.net/2012/08/16/dohbuh.gif ... frickin' MAC\nusers buy anything: http://www.ntk.net/2012/08/09/dohsx.gif\n... maybe not the best month for this branch of the W3C:\n\ngoto's considered non-harmful\n\nBoasting technical seminars, interplay of technological ideas\nand, of course, beer, this year's LINUXBIERWANDERUNG (from Sat\n2012-08-24, Doolin, Co. Clare, Ireland, for a \"small voluntary\ndonation\" per hiker) might just be the only walking holiday\nyou ever persuade your company to pay for. Continuing the\noutdoor theme, somewhat shorter notice for this Sunday's\nGREATER LONDON LINUX USER GROUP BLETCHLEY PARK PICNIC (around\n12noon, Sun 2012-08-18, Bletchley, near Milton Keynes, UKP6\nadmission fee). And for anyone who can't make the whole week\nin Ireland, there's some consolation in the form of LONDON\nPERLMONGERS PRESENT: DAMIEN CONWAY AT THE CONWAY HALL (6pm,\nTue 2012-08-27 and Thu 2012-08-29, Conway Hall, London WC1R,\nfree) - hopefully just the start of a similarly-named speaker/\nvenue trend which will also see Larry Wall performing at the\nGreat Wall Of China, and Eric Raymond at London's Raymond\nRevue Bar.\nhttp://www.lbw2012.draiocht.net/\n- sadly, less excuse for \"pingtime for Hitler\" gag this year\nhttp://list.ftech.net/pipermail/gllug/2012-August/025308.html\n- bring your \"Cryptonomicon\", please don't deface his Wiki\nhttp://london.pm.org/meetings/\n- last time: http://www.bluedust.com/projects/info7.asp\nhttp://www.gabba.net/pnp/lofi.asp\n- Plug'n'Play this Sun, feat. Casiotone & Thorsten Sideboard\nhttp://www.ica.org.uk/index.cfm?articleidv41\n- and Hacking @ ICA tonight, assuming you get this in time\n\nsufficiently advanced technology : the gathering\n\nDijkstra finally gone? Great. Now we can hang out with those\nsloppy little unrigorous languages he never let into the\nhouse. Right on cue: the latest dev version of GNU\nSMALLTALK, that dirty treat-me-like-an-object ducky. Version\n2.0b sprouts HTML-embedding and experimental MySQL support -\nimplying you might actually be able to do something useful\nwith it. Admittedly, \"useful\" has not been a word much used\nin the Smalltalk community recently. Many of the current\naddicts seem to live entirely in the pleasuredomes of that\nLand of Infinite Fun, Alan Kay's Smalltalk sequel SQUEAK,\nthe cross-platform UI-kablooie playground which, over time,\nbeen overloaded with more ill-fitting metaphors than this\nsentence. But even Squeak is showing signs of returning to\nplanet Earth. Talk of cleaning up the object tree are rife;\ncoders are working on XP-alike skins (instead of Squeak's\npeculiarly homely interface right now). There's even word of\na SmallTalkish scripting language in the offing. And all\nthose new MacOS X Objective C hackers are following the\ntrail back to the NeXT's great ancestor, giving a timely\nboost to the tribe. Xerox Parc *will* return!\nhttp://freshmeat.net/releases/93723/\n- the smaller Smalltalk\nhttp://www.squeak.org/download/\n- pretty colours\n- where they're all hiding\nhttp://www.iam.unibe.ch/~ducasse/WebPages/FreeBooks.html\n- look! free ST books!\nhttp://www.pocketsmalltalk.com/\n- look! free Palm IDE!\n\nceci n'est pas une http://www.gagpipe.com/\n\nwhere's Wendy?: http://3gamerchicks.com/features/bobsex/ ...\nhttp://news.bbc.co.uk/1/hi/2188214.stm got a bargain from:\nhttp://www.salon.com/tech/feature/2012/08/03/deleteddomains/print.html\n- still, makes a change from press releases from BT OPENWORLD:\nhttp://news.bbc.co.uk/1/hi/technology/2181078.stm ... life\nimitates http://www.thebrainstrust.co.uk/article.15.1961.html\n(from"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf", "copies": 3, "upstream_ids": ["20030228_hard_ham/hard_ham/00187.cc3d499cfabdc2d517c1ff46e9fa1fdf", "zefang-10930", "ealvaradob-10385"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 13296, "kept_chars": 5995}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00090-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-a66b40309f576c8c39bb", "state": {"channel": "email", "message": "Spring Special: Custom Banners and Flyers for Your Group\n\nDear Samantha,\n\nIs your local club or community group planning events this spring? Prairie Labs Print Services is offering discounted promotional materials for neighbourhood organisations.\n\nHere is what we can help you with:\n\n* Full-colour vinyl banners (up to 3 metres)\n* Double-sided glossy flyers in bulk\n* Branded tablecloths for registration desks\n* Custom t-shirts for volunteers\n\nAll orders placed by 15 February 2022 receive a 20% discount. Browse our full catalogue here:\n\nhttps://www.prairielabs.org/print-catalog\n\nWhether you are organising a fundraiser, a sports tournament or a street fair, professional signage makes a real difference.\n\nCall us at +1-312-555-0198 to discuss your project or place an order directly.\n\nWarm regards,\nRuth Chen\nOutreach Coordinator\nPrairie Labs Print Services\nprairielabs.org"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00090-2", "copies": 1, "upstream_ids": ["genu-00090-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-104:is_spam", "suite": "spam", "family": "sms_phishing-6fd84d6b71cfa831db62", "state": {"channel": "sms", "message": "Are you this much buzy .I don't think so"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-104", "copies": 1, "upstream_ids": ["mendeley-104"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-446:message_type", "suite": "spam", "family": "sms_phishing-ceb881a3502296f8b4f1", "state": {"channel": "sms", "message": "U should make a fb list"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-446", "copies": 2, "upstream_ids": ["mendeley-446", "uci-2971"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1:message_type", "suite": "spam", "family": "spamassassin-d05f75617a051416fd46", "state": {"channel": "email", "message": "Below is the result of your feedback form. It was submitted by\n(matthew_james@riversolutions.com) on Thursday, August 8, 2019 at 08:36:27\n\nmessage: - - - - - FREE FREE FREE! - - - - - Thats right free unlimited web cam feeds! - chat! - pictures! - videos! - ALL FREE!!! YOU PAY NOTHING!!! NOT ONE CENT! - This is a limited time offer because these gitls can't work for free so hurry up and click this link! http://teensforfree.cjb.net\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1", "copies": 3, "upstream_ids": ["20050311_spam_2/spam_2/01361.1b094e93a9a83d77859dcbf6637455d1", "zefang-12876", "ealvaradob-12187"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3816:is_spam", "suite": "spam", "family": "zefang_phishing_email-c82480d49d7008c30414", "state": {"channel": "email", "message": "win a free trip to antigua win a free trip to antigua just by visiting our site http : / / www . kingscasino . com anitgua 's largest casino and # 1 entertainment center now offers online gambling . open over 10 years with over 3 million visitors . king 's casino is a real casino ; it 's not just in cyberspace . located in downtown antigua where the cruise ships dock . so next time in antigua visit us in person during your cruise or from your hotel . or now on-line so click here at http : / / www . kingscasino . com featuring the finest graphics and the most fun to play ."}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3816", "copies": 2, "upstream_ids": ["zefang-3816", "ealvaradob-3661"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2520:message_type", "suite": "spam", "family": "sms_phishing-0d6da61938d4ec0ac007", "state": {"channel": "sms", "message": "Lol great now im getting hungry."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2520", "copies": 2, "upstream_ids": ["mendeley-2520", "uci-1424"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#549:message_type", "suite": "spam", "family": "nazario-756e412e136f4c839f4b", "state": {"channel": "email", "message": "(Chase) Message ID 169035 Client's Details Confirmation (Chase)\n\n0000-3748-PSI-0934\n\nDear valued Chase Bank Customer,\n\nDue to concerns, for the safety and integrity of the online banking community we\nhave issued the following warning message.\n\nIt has come to our attention that your account information needs to be confirmed\ndue to inactive customers, fraud and spoof reports. You have to renew your records in order to continue with the online service. However, failure to\nconfirm your records may result in your account suspension.\n\nOnce you have confirmed your account records your internet banking service will\nnot be interrupted and will continue as normal.\n\nTo confirm your bank account\nrecords please click here.\n\nThank you for your time,\n\nChase Online® Bank Billing Department."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#549", "copies": 2, "upstream_ids": ["phishing2.mbox#549", "phishing2.mbox#571"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-2517:is_spam", "suite": "spam", "family": "zefang_phishing_email-ae2b308c3465810e4ad1", "state": {"channel": "email", "message": "no doctor ' s visit . stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / imkjbest . com / z /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2517", "copies": 2, "upstream_ids": ["zefang-2517", "ealvaradob-2415"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-244:message_type", "suite": "spam", "family": "sms_phishing-5f920cddd8cd23c06f97", "state": {"channel": "sms", "message": "Hmm thinking lor..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-244", "copies": 2, "upstream_ids": ["mendeley-244", "uci-5244"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5065:is_spam", "suite": "spam", "family": "zefang_phishing_email-11aaf835f0727e9a949f", "state": {"channel": "email", "message": "invitation for you from rice university : the national forum on corporate finance mark , i left you a message regarding the national forum on corporate finance at rice . they would be delighted if you could serve as a panel member at this conference . here are the coordinates of the professor at rice who is in charge . i would appreciate if you could call him and let him know if you can attend . thanks . vince prof . david ikenberry jones graduate school of management rice university 713 - 348 - 5385"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5065", "copies": 2, "upstream_ids": ["zefang-5065", "ealvaradob-4851"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-2659:is_spam", "suite": "spam", "family": "zefang_phishing_email-3f0e6c945a62a3d07c41", "state": {"channel": "email", "message": "bulgarian summer classes in u . s . a . content - length : 1170 a colleague is interested in where intermediate bulgarian language classes might be offered this summer . first preference would be at a location in the united states . please reply directly to me : bret parker bparker @ uop . edu university of the pacific stockton , california 209-946 - 2014"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2659", "copies": 2, "upstream_ids": ["zefang-2659", "ealvaradob-2553"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-10617:is_spam", "suite": "spam", "family": "zefang_phishing_email-a3571adaac03ce71aa72", "state": {"channel": "email", "message": "enron management conference - home video project the enron management conference will be held wednesday through friday , november 14 - 16 , 2018 , at the westin la cantera resort in san antonio . while the business agenda is still being finalized , there is one thing that is certain . . . we want to inject some humor into the event . the past year has been eventful , and at times challenging , for enron and its employees . however , it ' s not in enron ' s corporate fabric to focus on the past . instead , we want to look ahead , and even laugh at ourselves . the management conference will provide a platform for just that . we want to challenge each of the business units to develop and submit a humorous \" home video \" to be presented at the management conference . the video can address any enron - related topic that your business unit chooses . however , there are a few rules to follow . 1 . ) the videos should be no more than five minutes long . 2 . ) business units should spend as little as possible on the videos , and no more than $ 10 , 000 . innovision communications ( beth stier ) has agreed to work with any business unit to help them create an original , inexpensive video . innovision can be contacted at 713 - 446 - 4424 . 3 . ) the videos should be fun , but not disparaging . they should adhere to enron ' s values of respect , integrity , communication and excellence . 4 . ) by monday , october 15 , business units need to notify me of their intention to submit a video . to avoid duplication , business units will also need to identify a topic they intend to address . 5 . ) to ensure inclusion in the management conference agenda , videos need to be submitted to me by friday , november 9 . 6 . ) awards will be given for creativity and humor . the first annual , and soon - to - be - coveted , \" enny \" will also be awarded for \" best overall video . \" if you have any questions about the \" home video \" project , please contact myself or mary clark . terrie james ph . 713 - 853 - 7727 terrie . james @ enron . com mary clark ph . 713 - 853 - 7325 mary . clark @ enron . com good luck and happy filming ! terrie james"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10617", "copies": 2, "upstream_ids": ["zefang-10617", "ealvaradob-10095"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4241:message_type", "suite": "spam", "family": "sms_phishing-f244bc3d606be2386ea3", "state": {"channel": "sms", "message": "Goodmorning, Today i am late for 5min."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4241", "copies": 1, "upstream_ids": ["mendeley-4241"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2605:is_spam", "suite": "spam", "family": "sms_phishing-aa4d73561a75a8af760f", "state": {"channel": "sms", "message": "What u talking bout early morning? It's almost noon where your at!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2605", "copies": 2, "upstream_ids": ["mendeley-2605", "uci-3062"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-488:is_spam", "suite": "spam", "family": "sms_phishing-cbb1f95591de818e060e", "state": {"channel": "sms", "message": "Ü eatin later but i'm eatin wif my frens now lei... Ü going home first?"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-488", "copies": 2, "upstream_ids": ["mendeley-488", "uci-1695"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5228:is_spam", "suite": "spam", "family": "sms_phishing-4bc84265cc718a1ca2dd", "state": {"channel": "sms", "message": "HEY BABE! FAR 2 SPUN-OUT 2 SPK AT DA MO... DEAD 2 DA WRLD. BEEN SLEEPING ON DA SOFA ALL DAY, HAD A COOL NYTHO, TX 4 FONIN HON, CALL 2MWEN IM BK FRMCLOUD 9! J X"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5228", "copies": 2, "upstream_ids": ["mendeley-5228", "uci-3527"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2139:message_type", "suite": "spam", "family": "sms_phishing-d876c6e8a76bee66f6bd", "state": {"channel": "sms", "message": "So can collect ur laptop?"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2139", "copies": 2, "upstream_ids": ["mendeley-2139", "uci-2147"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#343:is_spam", "suite": "spam", "family": "nazario-9e59c531d3fbedacc02e", "state": {"channel": "email", "message": "Mail Services\n\n?? ?? ?? ?? ?? >>> https://opaque-jade-germanium.glitch.me#lars.gray@willownetworks.com\n\n?? ?????? ??? ??? 24 ?? ??? ?????.\n\nRegards\n\n??? © 2017 ?? ??."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#343", "copies": 1, "upstream_ids": ["phishing-2023#343"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3467:message_type", "suite": "spam", "family": "sms_phishing-8f164392a516e5106daf", "state": {"channel": "sms", "message": "Thk some of em find wtc too far... Weiyi not goin... E rest i dunno yet... R ur goin 4 dinner den i might b able to join..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3467", "copies": 2, "upstream_ids": ["mendeley-3467", "uci-2819"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4629:is_spam", "suite": "spam", "family": "sms_phishing-ca1d857b07c958cdae0c", "state": {"channel": "sms", "message": "Tick, tick, tick ... Babe"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4629", "copies": 2, "upstream_ids": ["mendeley-4629", "uci-2745"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#44:is_spam", "suite": "spam", "family": "nazario-73667d00c8fe31d32235", "state": {"channel": "email", "message": "WE HAVE NOT RECEIVED YOUR ACCOUNT UPDATE - UPDATE NOW\n\nTo ensure delivery to your inbox, please add mbennett@web.de to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Feb 8, 2015. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 782 88\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#44", "copies": 1, "upstream_ids": ["phishing-2016#44"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#493:is_spam", "suite": "spam", "family": "nazario-8ee6e59dbc5db67f28ab", "state": {"channel": "email", "message": "E-mail Blacklisted\n\nDear sbyrne@yahoo.co.uk,\nOur current record indicates that your account was detected of spam activities and as a result has been blacklisted . Failure to confirm and update your account within 24 hours will lead to it being suspended permanently.\n\nReview and confirm your account below.\n\nReview/Confirm Your Email Address"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#493", "copies": 4, "upstream_ids": ["phishing-2016#493", "phishing-2018#172", "phishing-2019#215", "phishing-2020#131"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-16417:is_spam", "suite": "spam", "family": "zefang_phishing_email-8109e8a9c3dc14e12f8f", "state": {"channel": "email", "message": "start date : 1 / 4 / 02 ; hourahead hour : 1 ; start date : 1 / 4 / 02 ; hourahead hour : 1 ; no ancillary schedules awarded . variances detected . variances detected in load schedule . log messages : parsing file - - > > o : \\ portland \\ westdesk \\ california scheduling \\ iso final schedules \\ 2002010401 . txt - - - - load schedule - - - - $ $ $ variance found in table tblloads . details : ( hour : 1 / preferred : 365 . 91 / final : 365 . 88 ) trans _ type : final load _ id : pge 3 mkt _ type : 2 trans _ date : 1 / 4 / 02 sc _ id : enrj"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16417", "copies": 2, "upstream_ids": ["zefang-16417", "ealvaradob-15485"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12967:is_spam", "suite": "spam", "family": "zefang_phishing_email-3341c9ba0562fd43d663", "state": {"channel": "email", "message": "Tech Update Today\nVITAL SIGNS FOR JULY 12, 2012David Berlind\nWhat you need for iPaq wireless connectivity\nA public health organization has several iPaqs on order, but\nnow they want to add WAN connectivity to those PDAs. What\nexactly do they need to do? I'm so glad they asked. Here's everything you need to knowLatest from ZDNet News\nSecurity holds back Web services\nChina: Bursting with brainpower\nMicron cranks handle on next-gen memory\nJapan chipmakers team on 'SoCs' chip\nHP tech exec jumps ship to EMC\nHoles put Linux handheld at risk\nMicrosoft bids to be creme de la CRMMore Enterprise News\nFarber's PicksDanFarberDesperately seeking chief ethics officer\nGreedy CEOs are making a bad name for the profession--and corporate America. The title chief executive officer now\nconjures up an image of someone with questionable ethics, a\ngreedy disposition, and a disregard for shareholders. If\ngovernment-proposed remedies can\"t adequately address this\nethics crisis, disgruntled shareholders surely will.\nRead my full commentaryOracle challenges Microsoft anewOracle is taking on Microsoft's Exchange with a Collaboration Suite that includes e-mail, calendar and messaging software. Oracle has pitched its products as Microsoft replacements before -- with spectacularly little success. (At Comdex last November, the database giant offered its 9i and app-server software as a more secure alternative to Exchange.) What's different this time? Oracle is selling a separate collaboration product that, says CEO Larry Ellison, is cheaper than rival offerings. Plus: Companies facing a deadline to join Redmond's new licensing program may be more open to non-Micrososft alternatives.\nRead the full storyAlso: Ellison shrugs a No. 2 exec\nStandards stalled over royalty disputes\nThe World Wide Web Consortium is bracing for a vote next week that could decide once and for all how the standards body will handle patented technology that comes with royalties attached. If patented technology is rejected, the vote could mark a major shift in the Internet standards battle. Major intellectual property holders such as Microsoft and IBM have a lot at stake.\n\nRead the full storyMicrosoft's OpenGL claim shakes graphics world\nSurvey: Execs rethinking IT budgetsGrowing numbers of Fortune 1000 CIOs are re-evaluating their IT budgets and spending plans for coming months, according to a recent survey. Morgan Stanley finds that CIOs' top priorities include security software, application integration and Windows 2010/XP upgrades for the desktop. Integration challenges were slowing the rate of application adoption for 38 percent of those surveyed.Read more survey results\nOh look, a new bug in Outlook and IEA Danish security expert has warned of a security flaw in Microsoft's ActiveX scripting component that allows network attackers to peek at cookies, read files and even run programs on a victim's PC. The expert recommended that users disable ActiveX in IE and Outlook, at least until Microsoft releases a patch.Read the full story\nAlso: Outlook plug-in opens PCs to hackers\nWrite me at james47@yahoo.co.uk\nBack to topAlso on Tech Update Today\nCOMMENTARY Mixing Unix and WindowsMost enterprises have a mix of Windows and Unix systems. Larry Seltzer explores the options for interoperating those systems.\nQ&A Marcelo Tosatti: The future is Linux\nThe maintainer of the Linux kernel talks about new opportunities, defends UnitedLinux and tells why copyright laws keep him from entering the U.S.\nPREVIOUSLY ON TECH UPDATE TODAY ROI calculators don't tell the whole TCO story\nIn response to David Berlind's \"Dell's formula for ROI,\" TCO\nconsultant Tom Rankin warns that \"99 percent of TCO tools do\nnot look at the 'total' but may look at whatever parameters\nsuit the vendor to sell their products/services.\" Crucial Clicks--products worth looking at\nNetworkingFind the right home network Choosing the right technology to match your networking needs is vital. Here's help in making the choice.Read reviewsMost Popular ProductsNetworking\n1. Linksys EtherFast wireless AP\n2. Linksys EtherFast router\n3. Siemens SpeedStream router\n4. Wireless 802.11B Router\n5. NetGear HE102 802.11a wireless AP\nMore popular networking products\nElsewhere on ZDNetVeritas CEO Gary Bloom speaks out at Tech Update. Check out these personal laser printer picks at ZDNet Reviews. See how WAP works and how you should use it at Builder.com. Need a new job? Find one today in ZDNet's Career Center. CIOs talk out about the future of IT at CNET News.com.\nÂ\nÂ\n Manage\nMy Subscriptions | FAQ |\n Advertise\nHome |eBusiness | Security | Networking | Applications | Platforms | Hardware | Contact us\nCopyright 2012 CNET Networks, Inc. All rights reserved. ZDNet is a registered service mark of CNET Networks, Inc."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12967", "copies": 2, "upstream_ids": ["zefang-12967", "ealvaradob-12270"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-190:message_type", "suite": "spam", "family": "sms_phishing-a42650b6896338d7ad6e", "state": {"channel": "sms", "message": "hey, looks like I was wrong and one of the kappa guys numbers is still on my phone, if you want I can text him and see if he's around"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-190", "copies": 2, "upstream_ids": ["mendeley-190", "uci-4758"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3701:is_spam", "suite": "spam", "family": "sms_phishing-f3f6745ed20997959699", "state": {"channel": "sms", "message": "I am 6 ft. We will be a good combination!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3701", "copies": 2, "upstream_ids": ["mendeley-3701", "uci-2675"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-525:is_spam", "suite": "spam", "family": "sms_phishing-1004033ec0f127e8d6f0", "state": {"channel": "sms", "message": "Oi. Ami parchi na re. Kicchu kaaj korte iccha korche na. Phone ta tul na. Plz. Plz."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-525", "copies": 2, "upstream_ids": ["mendeley-525", "uci-1076"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00444-3:is_spam", "suite": "spam", "family": "generated_legit_email-dd62284fdd9e6f9d62f2", "state": {"channel": "email", "message": "Important updates to BeaconWorks community platform terms\n\nBeaconWorks Community Notice\n\nDear Rahul Bailey,\n\nWe are updating our Terms of Service for the BeaconWorks neighborhood platform. Please review the following changes carefully:\n\n- Data storage: All user profile data will now be hosted on servers within Germany to comply with local regulations.\n- Account verification: New members must verify their address using a utility bill before posting in buy-and-sell groups.\n- Content moderation: Automated filtering will be applied to all public posts starting 1 July 2015.\n- Dispute resolution: Any disagreements between users must first go through our internal mediation team.\n\nYou must accept these updated terms by 30 June 2015 to maintain access to your account. If you do not agree, your profile will be temporarily suspended until you confirm acceptance.\n\nTo review the full document, please log into your dashboard and navigate to Settings, then Legal Agreements.\n\nThank you for being part of our community.\n\nPierre Cox\nCommunity Manager\nBeaconWorks e.V."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00444-3", "copies": 1, "upstream_ids": ["gen-00444-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9:message_type", "suite": "spam", "family": "spamassassin-1ceeff01b5ba258e5592", "state": {"channel": "email", "message": "The MIME information you requested (last changed 3154 Feb 14)\n\nJustin Mason writes:\n\nI was puzzled at first, but I think I understand what happened.\n\nFirst, I approved the post because it didn't appear to be spam, even\nthough it wasn't from a member. I thought it was odd that someone\nwanted to send the MIME blurb to the list, but it was not really that\ndifferent from causing the New York Times web site to send a story to\nthe list. (Except that the bits here are antediluvian, but old bits are\na problem to be solved by social opporobrium, not technical constraints.)\n\nBut I think what actually happened is that some idiot got infected by\nKlez and had both FoRK and the pine-robot autoresponder address in their\nautoresponder, which responded. To FoRK."}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00344.99ffe210a9d425a86ab01840bdfd86c9", "zefang-7428", "ealvaradob-7117"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-1693:is_spam", "suite": "spam", "family": "zefang_phishing_email-103ecb972eeff7fdc12a", "state": {"channel": "email", "message": "next step defender extractions loafed caste particle fredrick blackboard lustiness hails adopted mounted explore resolver deadlines bivouacs nomination mcdougall waxing cosines ensnared invent tickle rainy accost whitened anarchist uncaught latinizations widener chesterton mohammedanize redeem broadens - - phone : 569 - 780 - 7589 mobile : 482 - 194 - 5436 email : berniecebenson @ kunmail . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1693", "copies": 2, "upstream_ids": ["zefang-1693", "ealvaradob-1626"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#124:message_type", "suite": "spam", "family": "nazario-ed9c0f49587a444645d0", "state": {"channel": "email", "message": "Message from eBay Member\n\neBay sent this message!\nYour registered name is included to show this message originated from eBay. Learn more.\n\nQuestion from eBay Member -- Respond Now\n\neBay sent this message on behalf of an eBay member via My Messages. Responses sent using email will not reach the eBay member. Use the Respond Now button below to respond to this message.\n\nQuestion from skaubraatmyra\n\nHello,\n\nHi,\n\nI signed up for Ebay and I've been an Ebay member for about 15 days. So, I'm carrying an \"icon\" where it says \"I am a member since less than 30 days\".\n\nA few days ago, I managed to sell an item. The guy who bought mine left me a positive feedback.\n\nBut still, mine's got a \"zero\" feedback?\n\nWhy is it that? If I click on my profile, I can see that he left me a feedback.\n\nBut the feedback counting doesn't go from zero to one.\n\nIs it because I'm a member who's been registered for less than 30 days?.\n\nRegards,\nGretta.\n\nRespond to this question in My Messages.\n\nThank you for using eBay\n\nhttp://www.ebay.com/\n\nMarketplace Safety Tip\n\nIf this message is an offer to sell an item without winning it on the eBay Web site (including Second Chance Offers sent through My Messages) please do not respond to the sender. These external transactions are unsafe and not covered by eBay purchase protection programmes.\n\nNever pay for your eBay item through instant wire transfer services such as Western Union or MoneyGram. These payment methods are unsafe when paying someone you do not know.\n\nIs this email inappropriate? Does it breach eBay policy? Help protect the community by reporting it.\n\nThis email appears in the language of the eBay site where you are registered.\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.co.uk/education/spooftutorial\n\nThis eBay notice was sent to christopher_perez@live.com on behalf of another eBay member through the eBay platform and in accordance with our Privacy Policy. If you would like to receive this email in text format, change your notification preferences.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's communication policies.\nPrivacy Policy: http://pages.ebay.co.uk/help/policies/privacy-policy.html\nUser Agreement: http://pages.ebay.co.uk/help/policies/user-agreement.html\n\nCopyright © 2013 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#124", "copies": 3, "upstream_ids": ["phishing2.mbox#124", "phishing2.mbox#138", "phishing2.mbox#141"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00553-1:message_type", "suite": "spam", "family": "generated_legit_email-5c63b6edd98c2aeac99e", "state": {"channel": "email", "message": "Professional Liability Insurance Renewal Notice\n\nHi Eric,\n\nI wanted to follow up regarding our professional liability insurance renewal. The current policy held under UnionWorks GmbH is set to expire on December 15, 2019, so we need to finalize the paperwork before then to avoid any lapse in coverage.\n\nThe premium for the upcoming year will be €3,240.00, which reflects a slight adjustment from last term. I have already uploaded the draft renewal agreement to our internal portal for your review. Please log in and read through the updated terms when you have a moment. You can access the document here: https://www.unionworks.de/insurance/renewal-2019\n\nWe must submit the signed renewal form by December 1, 2019 at the latest. If you notice any discrepancies in the coverage limits or have questions about the new clauses, just let me know and I will arrange a call with the broker.\n\nThanks for handling this promptly.\n\nBest regards,\nKwame Muller\nOperations Director\nUnionWorks GmbH"}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00553-1", "copies": 1, "upstream_ids": ["gen-00553-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:sms_phishing:none:mendeley-3852:message_type", "suite": "spam", "family": "sms_phishing-69276f7aa7c231be9408", "state": {"channel": "sms", "message": "Aight sorry I take ten years to shower. What's the plan?"}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3852", "copies": 2, "upstream_ids": ["mendeley-3852", "uci-1339"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2801:is_spam", "suite": "spam", "family": "sms_phishing-6135d6422aec7db149fd", "state": {"channel": "sms", "message": "Download as many ringtones as u like no restrictions, 1000s 2 choose. U can even send 2 yr buddys. Txt Sir to 80082 £3"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2801", "copies": 2, "upstream_ids": ["mendeley-2801", "uci-2352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2082:message_type", "suite": "spam", "family": "sms_phishing-0ac0fca1ece4f358ef5f", "state": {"channel": "sms", "message": "Thought praps you meant another one. Goodo! I'll look tomorrow"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2082", "copies": 2, "upstream_ids": ["mendeley-2082", "uci-1723"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-69:message_type", "suite": "spam", "family": "sms_phishing-d4df68cfb848ffe87070", "state": {"channel": "sms", "message": "You need to get up. Now."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-69", "copies": 2, "upstream_ids": ["mendeley-69", "uci-3936"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00256-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-09d03cc15b8a8096d158", "state": {"channel": "email", "message": "Action Required: Customs Clearance Fee for Delivery\n\nDear Anna Nguyen,\n\nWe are contacting you regarding a parcel currently held at our regional distribution centre. The package could not be delivered because a small customs clearance fee remains outstanding. As Atlas Works Financial Services manages the escrow payments for this shipment, we have been instructed to collect the balance before the courier can proceed.\n\nPlease review the shipment details below and settle the fee so that your delivery can be released without further delay.\n\nRecipient: Anna Nguyen\nEmail on file: dennis.cooper@aol.com\nTracking Reference: AW-992810447\nCarrier: Atlas Express Logistics\nDelivery Attempt Date: 25 April 2008\nStatus: Held at Customs\nAmount Due: GBP 3.45\nPayment Deadline: 30 April 2008\n\nTo ensure the security of your transaction, the payment must be made using a valid debit or credit card through our encrypted portal. Once the fee has been successfully processed, your parcel will be dispatched for redelivery within twenty-four hours.\n\nPlease click the link below to verify your card details and pay the outstanding amount:\nhttps://www.atlasworks-secure.co.uk/payments/customs-clearance/AW-992810447\n\nIf the fee is not settled by the deadline shown above, the parcel will be returned to the sender, and additional return shipping charges may apply. We understand that unexpected fees can be inconvenient, but this charge is required by international shipping regulations for goods entering the country.\n\nShould you experience any difficulty completing the payment online, please reply to this message and our support team will be happy to assist you.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nGary James\nPayment Resolution Team\nAtlas Works Financial Services"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00256-3", "copies": 1, "upstream_ids": ["genu-00256-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-7617:is_spam", "suite": "spam", "family": "zefang_phishing_email-8f8b57649b58c9d273e8", "state": {"channel": "email", "message": "discussions louise : i have attached two \" job descriptions \" based on my conversations with dave , john , tom m . and you . the first focuses on power / gas origination in texas / mexico as a seperate region , the second focuses on power asset development business in north america . i know john sees a need to focus on re - establishing our position in the texas gas maket . if we combine the ercot / mexico power origination with texas and mexico gas , then this will build on my power skills , knowledge of the texas gas market , and my non - us market experience . dave is on board with me heading up the power development activities and probably would be more on board with me focusing on ercot as this would be less disruptive to his business . i am open to either - we can chat tomorrow . brian"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7617", "copies": 2, "upstream_ids": ["zefang-7617", "ealvaradob-7292"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00832-5:is_spam", "suite": "spam", "family": "generated_legit_email-76fc85773d99f218164b", "state": {"channel": "email", "message": "Your commercial liability policy renewal is due\n\nHi Sofia,\n\nJust a quick note to remind you that your commercial liability insurance policy with Cedar Solutions (Policy #CS-4491) expires on 30 April 2014.\n\nPlease log in to our portal at https://www.cedarsolutions.org/renewals to review your updated quote and complete the renewal process before the deadline to avoid any lapse in coverage.\n\nLet me know if you have questions.\n\nBest,\nRonald Ramirez\nCedar Solutions Insurance"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00832-5", "copies": 1, "upstream_ids": ["gen-00832-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:nazario:none:phishing-2025#89:message_type", "suite": "spam", "family": "nazario-9bad123c04f88c371b4b", "state": {"channel": "email", "message": "Notification: Expiry Alert 3/10/2008\n\nrahul.silva@blueworks.biz,\n\nThis is to notify you Your password for rahul.silva@blueworks.biz expires today.\nFollow instruction below to keep your current password and update your account..\n\nKeep Current Password https://b2hu3hhssz.eu-west-3.awsapprunner.com/?username=rahul.silva@blueworks.biz\n\n© 2008\n\nPrivacy Policy https://ermt4sdhyk.us-east-1.awsapprunner.com/?username=sales@crestbadges.com\n\n**DISCLAIMER:** This message is confidential and intended solely for the recipient. If you have received this message in error, please notify the sender immediately and delete it from your system. No confidentiality or privilege is waived or lost due to any mis-transmission."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#89", "copies": 1, "upstream_ids": ["phishing-2025#89"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00672-1:message_type", "suite": "spam", "family": "generated_legit_email-3e8547f4a6f218771fe4", "state": {"channel": "email", "message": "Your November statement is ready\n\nHi Tomasz,\n\nThanks for reaching out earlier. I have checked your account and can confirm that your monthly QuartzLabs credit card statement for October 2022 is now available to view online. The total balance due is £412.55, with a minimum payment of £25.00 required by 20 November 2022.\n\nYou can log in and download your statement securely here: https://www.quartzlabs.de/statements/nov2022\n\nLet me know if you need any further assistance with this.\n\nBest regards,\nJulie Walsh\nQuartzLabs Customer Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00672-1", "copies": 1, "upstream_ids": ["gen-00672-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:generated_legit_email:none:gen-00807-5:message_type", "suite": "spam", "family": "generated_legit_email-85297848c1b4e236108d", "state": {"channel": "email", "message": "Neighbourhood cleanup this weekend\n\nHi Amy,\n\nYes, the Ridgewood Community Cleanup is confirmed for Saturday 23 May at 9 AM. We are meeting by the main park entrance. Bring gloves if you have them!\n\nRSVP here: https://www.ridgesystems.de/events/cleanup2015\n\nCall me on 07700 900461 if you get lost.\n\nAshley Shah"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00807-5", "copies": 1, "upstream_ids": ["gen-00807-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:sms_phishing:none:mendeley-2857:is_spam", "suite": "spam", "family": "sms_phishing-91f973a250a9b1931f07", "state": {"channel": "sms", "message": "Thanks for your Ringtone Order, Reference T91. You will be charged GBP 4 per week. You can unsubscribe at anytime by calling customer services on 09057039994"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2857", "copies": 3, "upstream_ids": ["mendeley-2857", "mendeley-4216", "uci-375"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00550-3:message_type", "suite": "spam", "family": "generated_legit_email-ba69353288d8de7bc60e", "state": {"channel": "email", "message": "Donation Receipt - Meridian Labs Annual Fund\n\nMeridian Labs Charitable Trust\nDonation Confirmation\n\nDonor: Kenneth Jackson\nEmail: tomasz.khan@mail.com\nReceipt Number: ML-2015-03-4491\nAmount: $250.00\nFund: Annual Research Endowment\nPayment Method: Visa ending in 4021\n\nDear Kenneth,\n\nThank you for your generous contribution to the Meridian Labs Annual Research Endowment. Your donation of $250.00 has been successfully processed. This receipt serves as confirmation for your tax records. Meridian Labs is a registered 501(c)(3) organization, and your gift is tax-deductible to the extent allowed by law.\n\nTo view your complete donation history or download a PDF copy of this receipt, please visit https://www.meridianlabs.org/donor-receipts.\n\nWe deeply appreciate your continued support of our research initiatives.\n\nSincerely,\nRahul Scott\nDirector of Development\nMeridian Labs Charitable Trust"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00550-3", "copies": 1, "upstream_ids": ["gen-00550-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297:message_type", "suite": "spam", "family": "spamassassin-285b7c334e4aa8889c29", "state": {"channel": "email", "message": "CD player UI for toddlers\n\nthe\nOr some sound indicating that the music will start real soon now.\n\nyou\nIdempotency everywhere you look...\n\nIt'd be an MP3 player with solid state storage... instant on."}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1076:message_type", "suite": "spam", "family": "sms_phishing-1a504742bb3cac07b029", "state": {"channel": "sms", "message": "I like cheap! But i‘m happy to splash out on the wine if it makes you feel better.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1076", "copies": 2, "upstream_ids": ["mendeley-1076", "uci-5199"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a:is_spam", "suite": "spam", "family": "spamassassin-dd3c741bcb76dff5d15a", "state": {"channel": "email", "message": "Headlines for 2024-08-09\n\nuse Perl Daily Headline Mailer\n\nDamian Conway to speak in London\nposted by ziggy on Thursday August 08, @15:21 (damian)\nhttp://use.perl.org/article.pl?sid\u0002/08/08/1923211\n\nMeeting in Budapest\nposted by ziggy on Thursday August 08, @17:24 (groups)\nhttp://use.perl.org/article.pl?sid\u0002/08/08/1927253\n\nCopyright 2019-2024 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3951:message_type", "suite": "spam", "family": "sms_phishing-5fd86e1563095dab07e3", "state": {"channel": "sms", "message": "I meant as an apology from me for texting you to get me drugs at &lt;#&gt; at night"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3951", "copies": 2, "upstream_ids": ["mendeley-3951", "uci-3621"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3417:is_spam", "suite": "spam", "family": "zefang_phishing_email-02eb1664de1b68454abe", "state": {"channel": "email", "message": "i have you scheduled . dolores vince j kaminski 08 / 30 / 2011 08 : 13 am to : dolores muzzy / hou / ect @ ect cc : shirley crenshaw / hou / ect @ ect , vince j kaminski / hou / ect @ ect subject : re : dolores please , register me for a session on 9 / 29 / 2011 at 12 : 45 . vince kaminski celeste roberts 08 / 29 / 2011 06 : 21 pm to : celeste roberts / hou / ect @ ect cc : ( bcc : vince j kaminski / hou / ect ) subject : urgent the associate and analyst recruiting department will be conducting a number of two hour workshops to review our recruiting and interview process for the fall on - campus recruiting effort . critical information regarding our on - campus interview process , revised evaluation forms and program structure will be reviewed during these two hours sessions . it is mandatory that all team members attend these workshops . all team members must attend in order to articulate and demonstrate the enron recruiting process . knowing how busy schedules are , we have made arrangements to present these workshops in two hours sessions for a total of 40 workshops that will run during the last week of august , through the month of september and end at mid october . listed below are the dates , location and times for each session . please select a date and time and e - mail this information to my assistant , dolores muzzy . we can accommodate 25 participants at a time . dolores will take dates and times on a first come , first serve basis . we have scheduled enough sessions to accommodate every member of both the associate and analyst recruiting teams . in order to participate in the recruiting process , you must attend one of these sessions . we will be tracking participation . cpe credits will also be given for attending this workshop ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3417", "copies": 2, "upstream_ids": ["zefang-3417", "ealvaradob-3279"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00394-4:message_type", "suite": "spam", "family": "generated_legit_email-9bf4e222e7a7f1f2d31e", "state": {"channel": "email", "message": "Your package is on the way!\n\nHey Rahul,\n\nHope you are doing well! I finally got around to shipping those books and the extra monitor I promised you last month. The box left my place yesterday afternoon and should be arriving soon. I wanted to give you all the details so you know what to look out for.\n\nHere is everything you need to know about the shipment:\n\n- Carrier: Union Solutions Express\n- Tracking number: USX-9938271044-B\n- Package contents: Three hardcover books (the architecture ones you wanted) and the 24-inch Dell monitor\n- Estimated delivery date: Thursday, 24 September 2009\n\nA few quick steps to make sure everything goes smoothly when it arrives:\n\n1. Keep an eye out for the delivery truck on Thursday. They usually drop off packages at the front desk of your building between 1 PM and 4 PM.\n2. Check the outside of the box for any visible damage before you sign for it. The monitor was heavy, so I wrapped it in a ton of bubble wrap, but carriers can be rough sometimes.\n3. If you are not going to be at the office on Thursday, please let your receptionist know to accept it on your behalf by Wednesday, 23 September 2009 at the latest. I do not want it sitting out in the rain or getting returned to sender.\n4. Once you open it up, test the monitor right away. If the screen has any dead pixels or the cables I included do not work, tell me immediately so I can file a claim with the carrier before the deadline passes.\n\nI am really hoping it gets there safely and in time for your new desk setup. Let me know as soon as you have unboxed it! We should catch up properly over a call sometime next week once you are settled.\n\nTake care,\nIvan Morgan"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00394-4", "copies": 1, "upstream_ids": ["gen-00394-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:generated_legit_email:none:gen-00897-3:message_type", "suite": "spam", "family": "generated_legit_email-9cc85e22e6bdebe98158", "state": {"channel": "email", "message": "Marketplace inquiry regarding the Dell monitor listing\n\nHi Laura,\n\nI am following up on the internal marketplace listing you posted last week for the Dell 27-inch monitor. A colleague from the design team is interested in purchasing it for their home office setup, and I wanted to coordinate the details with you before the end of the week.\n\nPlease review the following steps so we can complete the transfer smoothly:\n\n1. Confirm the current condition of the monitor and whether you still have the original power cable and stand included.\n2. Update the asking price if necessary; the buyer mentioned they are comfortable paying up to $185.00.\n3. Arrange a time for the handover. The buyer is available on Tuesday 8 January or Wednesday 9 January between 1:00 PM and 4:00 PM in the main lobby.\n4. Once the exchange is done, mark the item as sold on our internal portal here: https://www.deltalabs.co.uk/marketplace/listings\n5. Forward the confirmation email to me so I can close out the ticket in our facilities system.\n\nThe buyer also asked if you could wipe any saved display profiles from the monitor settings before handing it over. They want to ensure it connects easily to their Mac laptop without any driver conflicts.\n\nIf you run into any issues accessing the marketplace link above, let me know and I can walk you through it. I will be in the office all day tomorrow if you prefer to discuss this in person.\n\nThanks for helping us keep equipment circulating within the company rather than ordering new stock. It really helps our quarterly budget.\n\nBest regards,\nCarlos Kim\nFacilities Manager\nDelta Labs\nPhone: (555) 209-4417"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00897-3", "copies": 1, "upstream_ids": ["gen-00897-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:generated_unwanted_email:none:genu-00278-1:message_type", "suite": "spam", "family": "generated_unwanted_email-7077bc0a4d6f29e43a4e", "state": {"channel": "email", "message": "Important Update: Change of Bank Details for Ridge Foods\n\nRIDGE FOODS - INVOICE & PAYMENT UPDATE\n\nDear Julie Garcia,\n\nThank you for your continued partnership with Ridge Foods. We are writing to inform you of an important administrative change regarding our payment processing.\n\nDue to a recent transition to a new corporate banking partner, our bank account details have changed effective immediately. Please ensure that all future payments, including the outstanding invoice #RF-2010-4492 currently due from Union Solutions, are directed to our new account.\n\nNew Banking Details:\nBank Name: Global Commerce Bank\nAccount Name: Ridge Foods International Ltd\nSort Code: 40-22-18\nAccount Number: 88391022\nIBAN: GB42 GCOB 4022 1888 3910 22\nSWIFT/BIC: GCOBGB2L\n\nPlease update your accounting records accordingly before processing the next scheduled transfer on 10 May 2010. Payments sent to the previous account will be delayed and may incur return fees.\n\nTo review your current statement and verify these changes in our portal, please use the button below:\n\nUpdate Payment Details\n\nhttps://www.ridgefoods42.net/vendor-portal/update-payment-details\n\nIf you have any questions or require verbal confirmation before updating your systems, please do not hesitate to contact our accounts receivable department at +1 800 555 0194.\n\nWe appreciate your prompt attention to this matter and look forward to continuing our successful business relationship.\n\nBest regards,\nRonald Chen\nSenior Accounts Manager\nRidge Foods\n\nRidge Foods International Ltd\n450 Industrial Parkway, Suite 200\nChicago, IL 60614, United States\nPhone: +1 800 555 0194\nEmail: accounts@ridgefoods.com\n\nPrivacy Policy: https://www.ridgefoods.com/privacy\nConfidentiality Notice: This email and any attachments are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error, please notify the sender immediately and delete this message from your system."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00278-1", "copies": 1, "upstream_ids": ["genu-00278-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4071:message_type", "suite": "spam", "family": "sms_phishing-f767cd50eb64bcf7ad57", "state": {"channel": "sms", "message": "Take us out shopping and Mark will distract Isaiah.=D"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4071", "copies": 2, "upstream_ids": ["mendeley-4071", "uci-3076"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3603:is_spam", "suite": "spam", "family": "sms_phishing-17ae2cc4028c54544f27", "state": {"channel": "sms", "message": "Pls clarify back if an open return ticket that i have can be preponed for me to go back to kerala."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3603", "copies": 2, "upstream_ids": ["mendeley-3603", "uci-5040"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00146-3:message_type", "suite": "spam", "family": "generated_legit_email-25dc6cf16795dfe1a90a", "state": {"channel": "email", "message": "Direct Debit Confirmation – November Invoice\n\nHello Catherine Ross,\n\nThis email confirms that your direct debit payment has been successfully processed by Willow Group Services.\n\nPayment Details:\n- Invoice Number: WG-2018-8841\n- Amount: £124.50\n- Date Processed: 21 November 2018\n- Account Reference: SR-99201\n- Status: Paid\n\nNext Steps:\n1. Your account balance is now fully up to date.\n2. Please review your digital receipt by logging in at https://www.willowgroup.net/billing/receipts.\n3. Ensure your linked bank account remains active before the next billing cycle on 21 December 2018 to avoid any late fees.\n\nIf you did not authorise this transaction or notice an error, you must contact us no later than 28 November 2018. Reach out to Daniel Muller in our billing department at 020 7946 0123 for immediate assistance.\n\nThank you for your continued business.\n\nDaniel Muller\nAccounts Receivable\nWillow Group Services"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00146-3", "copies": 1, "upstream_ids": ["gen-00146-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:generated_unwanted_email:none:genu-00132-0:message_type", "suite": "spam", "family": "generated_unwanted_email-13fb02127f38cf4c9e2b", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected on your Orbit Design account\n\nDear Carlos Garcia,\n\nWe detected an unusual sign-in to your Orbit Design account from São Paulo, Brazil. To protect your information, we have temporarily locked your account.\n\nIf this was not you, please complete the following steps by 23 July 2013:\n\n1. Visit our secure verification page at https://www.myorbitdesign-online.com/verify-identity\n2. Confirm your personal details and reset your password.\n3. Review your recent activity log.\n\nFailure to verify your identity by the deadline may result in permanent suspension of your account.\n\nThank you for helping us keep your workspace secure.\n\nKind regards,\nElizabeth Young\nAccount Security Team\nOrbit Design\nsupport@orbitdesign.io"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00132-0", "copies": 1, "upstream_ids": ["genu-00132-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2356:message_type", "suite": "spam", "family": "sms_phishing-796fe29cd899d191fa69", "state": {"channel": "sms", "message": "Dear Voucher Holder, 2 claim this weeks offer, at your PC go to http://www.e-tlp.co.uk/expressoffer Ts&Cs apply.2 stop texts txt STOP to 80062."}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2356", "copies": 2, "upstream_ids": ["mendeley-2356", "uci-635"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15635:is_spam", "suite": "spam", "family": "zefang_phishing_email-0a3002dff679a54a2586", "state": {"channel": "email", "message": "operations assignments welcome back , again . you will have to let me know when you have baby pictures in the office - would love to see them . last week , i orchestrated some changes within my operations organization which will free up more of my time to focus on the successful commercialization of mid and back office services . i have asked bob hall to lead operations for enron americas , with peggy hedstrom ( calgary operations ) , jeff gossett ( gas risk and south america operations ) , stacey white ( power risk ) and leslie reeves ( confirmations and settlements ) reporting directly to him . he is enthusiastic about the additional responsibility and i believe that he will do an excellent job for enron americas . other changes include appointing kevin sweeney as head of operations for eim , with brenda herod reassigned as part of a project team that will work on the ees assimilation . i have also asked beth apollo to play a project management lead role in the assimilation of ees into ews . i will be glad to fill you in on more details on any of these changes . the result is to give more responsibility to several key players - stretching them , but i will be there to support and coach them in their new roles if needed . plus , i can now devote significantly more time to the commercialization effort . in that regard , i did move to the 27 th floor ( next to piper ) while you were out on leave . i wanted to share these changes with you before a memo goes out across the organization next monday . call me if you would like to discuss . - - sally"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15635", "copies": 2, "upstream_ids": ["zefang-15635", "ealvaradob-14777"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00722-4:is_spam", "suite": "spam", "family": "generated_legit_email-6c14167e39ed146af6f0", "state": {"channel": "email", "message": "Lakeside Logistics Charity Drive Receipt\n\nLAKESIDE LOGISTICS\n\nHi Ling Li,\n\nThank you for your $50 donation to our annual charity drive. Please submit your matching pledge form to HR by February 29, 2010 to double your impact.\n\nView Your Receipt\n\nLakeside Logistics, Inc.\n88 Harbor Blvd, Duluth, MN 55802\nPrivacy: lakesidelogistics.net/privacy\n\nArjun Moore\nOperations Manager"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00722-4", "copies": 1, "upstream_ids": ["gen-00722-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:nazario:none:phishing-2023#383:is_spam", "suite": "spam", "family": "nazario-6185a6e86e0efe6b1e78", "state": {"channel": "email", "message": "Your summitfoods.org: Password Expire Request = 07\n\nDear richard,\n\nCurrent password for (helen.schmidt@granitenetworks.biz) expires today: EST. 11/22/2009\n\nTo keep Current password check below\n\nKeep Same Password http://klvt.avuidf.difarmafarmaceutica.com.br/#\n\nAdministrator for emberdesign.io Account and services.\n© 2009"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#383", "copies": 1, "upstream_ids": ["phishing-2023#383"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4716:is_spam", "suite": "spam", "family": "sms_phishing-c821f21bc2793dbf7814", "state": {"channel": "sms", "message": "Oh k :)why you got job then whats up?"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4716", "copies": 2, "upstream_ids": ["mendeley-4716", "uci-2178"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2960:is_spam", "suite": "spam", "family": "sms_phishing-dcda6ea42d1ad09aded8", "state": {"channel": "sms", "message": "Hi:)did you asked to waheeda fathima about leave?"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2960", "copies": 2, "upstream_ids": ["mendeley-2960", "uci-4345"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2313:message_type", "suite": "spam", "family": "sms_phishing-f8b0e97c7095b010cba2", "state": {"channel": "sms", "message": "Nope watching tv at home... Not going out. V bored..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2313", "copies": 2, "upstream_ids": ["mendeley-2313", "uci-1019"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_spam:none:uci-3976:is_spam", "suite": "spam", "family": "sms_spam-e3d4765a03520a66a1f0", "state": {"channel": "sms", "message": "do u think that any girl will propose u today by seing ur bloody funky shit fucking face...............asssssholeeee................"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_spam", "split": null, "upstream_id": "uci-3976", "copies": 1, "upstream_ids": ["uci-3976"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3637:is_spam", "suite": "spam", "family": "sms_phishing-df4aa3f71d123301f541", "state": {"channel": "sms", "message": "Wat time r ü going to xin's hostel?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3637", "copies": 2, "upstream_ids": ["mendeley-3637", "uci-4733"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6608:is_spam", "suite": "spam", "family": "zefang_phishing_email-204a57ec04dd9339cc98", "state": {"channel": "email", "message": "wccfl 2015 call for papers wccfl xvii west coast conference on formal linguistics university of british columbia february 20-22 , 2015 call for papers abstracts are invited for 20 - minute talks in all areas of linguistics from any theoretical perspective . there will be two special sessions in addition to the main program . abstracts may be submitted to the main program or either of the special sessions . * special sessions : interfaces * invited speakers will be presenting at the following 2 special sessions : lisa selkirk , umass : phonology / syntax interface angelika kratzer , umass : syntax / semantics interface abstract requirements abstracts should be no more than one standard size page in length with the option of including an additional page for data and references . abstracts should be in at least 11 - point type with 1 - inch margins , single-spaced . please mail ten anonymous copies as well as a 3 \" x 5 \" index card with the following information : title name of author ( s ) mailing address affiliation ( s ) area ( phonology , syntax , semantics . . . ) phone number ( optional ) e-mail address ( optional ) please specify on your index card if you are submitting your abstract to either of the special sessions . submissions are limited to 1 individual and 1 joint abstract per author . please do not send abstracts by e-mail . * abstracts must be received by november 14 , 2014 * abstracts should be sent to : wccfl xvii committee department of linguistics university of british columbia e270-1866 main mall vancouver , british columbia canada v6t 1z1 information will be made available at the wccfl web site : http : / / www . interchange . ubc . ca / msr / wccfl . html /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6608", "copies": 2, "upstream_ids": ["zefang-6608", "ealvaradob-6332"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#493:message_type", "suite": "spam", "family": "nazario-8ee6e59dbc5db67f28ab", "state": {"channel": "email", "message": "E-mail Blacklisted\n\nDear sbyrne@yahoo.co.uk,\nOur current record indicates that your account was detected of spam activities and as a result has been blacklisted . Failure to confirm and update your account within 24 hours will lead to it being suspended permanently.\n\nReview and confirm your account below.\n\nReview/Confirm Your Email Address"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#493", "copies": 4, "upstream_ids": ["phishing-2016#493", "phishing-2018#172", "phishing-2019#215", "phishing-2020#131"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00023-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-f5a021afc1d58fa36d32", "state": {"channel": "email", "message": "Exclusive Summer Clearance Sale\n\nEmberTech Summer Clearance\n\nDear Raymond Bell,\n\nEnjoy up to 60% off select laptops, monitors, and office furniture this August.\n\nShop the Sale Now\nhttps://www.embertech.biz/clearance-2012\n\nThank you for shopping with EmberTech.\nSophie Brown\nSales Department\n\nEmberTech Ltd, 400 Industrial Way, Austin, TX 78701\nPrivacy Policy: https://www.embertech.biz/privacy"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00023-4", "copies": 1, "upstream_ids": ["genu-00023-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk discount offer on electronics, clothes or furniture from an unknown company", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00007-1:is_spam", "suite": "spam", "family": "generated_legit_email-c4413d64a2b4cd6f7670", "state": {"channel": "email", "message": "Action required: Q2 Compliance Review documents\n\nDear Steven,\n\nAs part of our quarterly compliance process, Crest Consulting requires your team to submit the updated review documents by Friday 27 May 2010. Please follow the steps below:\n\n1. Download the current Q2 checklist from https://www.crestconsulting.io/compliance/q2-checklist-2010.\n2. Update sections 3 and 4 with your department's latest figures.\n3. Have your line manager sign off on page 6.\n4. Email the completed PDF back to me no later than 17:00 BST on 27 May.\n\nIf any data is missing or you anticipate delays, please flag it immediately so we can adjust the schedule. Late submissions will delay the entire audit cycle.\n\nThank you for your prompt attention to this matter.\n\nBest regards,\nBenjamin Carter\nCompliance Manager\nCrest Consulting\ndeborah_hughes@aol.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00007-1", "copies": 1, "upstream_ids": ["gen-00007-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-370:is_spam", "suite": "spam", "family": "sms_phishing-2f746625a22893de654e", "state": {"channel": "sms", "message": "No one interested. May be some business plan."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-370", "copies": 2, "upstream_ids": ["mendeley-370", "uci-5352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-296:message_type", "suite": "spam", "family": "sms_phishing-0f21f2f04022113b4acd", "state": {"channel": "sms", "message": "Pls dont forget to study"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-296", "copies": 2, "upstream_ids": ["mendeley-296", "uci-4262"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00743-3:is_spam", "suite": "spam", "family": "generated_legit_email-0afff79ee2ba2f8c82e5", "state": {"channel": "email", "message": "Equipment shipment dispatched - tracking enclosed\n\nFalcon Media Dispatch Notice\n\nHi Sophie,\n\nThe replacement monitors were shipped today. Please confirm receipt by August 28, 2008.\n\nTracking Number: FM-99281-2008\n\nTrack your package\nhttps://www.falconmedia.com/tracking/FM-99281-2008\n\nRegards,\nOlivia Chen\nIT Procurement\n\nFalcon Media Inc., 400 Broad St, Seattle, WA 98109\nPrivacy Policy: https://www.falconmedia.com/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00743-3", "copies": 1, "upstream_ids": ["gen-00743-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-3467:is_spam", "suite": "spam", "family": "sms_phishing-8f164392a516e5106daf", "state": {"channel": "sms", "message": "Thk some of em find wtc too far... Weiyi not goin... E rest i dunno yet... R ur goin 4 dinner den i might b able to join..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3467", "copies": 2, "upstream_ids": ["mendeley-3467", "uci-2819"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4944:is_spam", "suite": "spam", "family": "zefang_phishing_email-a6798c31dc6b9788c620", "state": {"channel": "email", "message": "desk to desk good afternoon mr . williams your deal number for the swap is 646795 for a price of $ 56 . 15 . kysa m . alport enron north america 503 - 464 - 7486"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4944", "copies": 2, "upstream_ids": ["zefang-4944", "ealvaradob-4733"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#246:is_spam", "suite": "spam", "family": "nazario-f681634c617f1829189c", "state": {"channel": "email", "message": "System Administrator\n\nDue to technical reasons, please we are expanding and upgrading all Staff Mailbox immediately. Please CLICK HERE <http://site9214306.92.webydo.com/?v=1> and fill the form completely. Click on the send button for validation. This message is from Admin Helpdesk.\n\nThe views expressed in this email are, unless otherwise stated, those of the author and not those of the National Health Laboratory Service or its management. The information in this e-mail is confidential and is intended solely for the addressee. Access to this e-mail by anyone else is unauthorized. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted in reliance on this, is prohibited and may be unlawful. Whilst all reasonable steps are taken to ensure the accuracy and integrity of information and data transmitted electronically and to preserve the confidentiality thereof, no liability or responsibility whatsoever is accepted if information or data is, for whatever reason, corrupted or does not reach its intended destination."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#246", "copies": 1, "upstream_ids": ["phishing-2015#246"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00010-1:message_type", "suite": "spam", "family": "generated_unwanted_email-c6c42eb256a655088ade", "state": {"channel": "email", "message": "Urgent: Mailbox Storage Quota Exceeded – Immediate Action Required\n\nNotice of Mailbox Quota Limit\n\nRecipient: Andrew Weber (rmoore@yahoo.com)\nMailbox Status: Full (100% Capacity)\nReference Number: IT-2010-0208-AW\nDepartment: Information Technology Services, Oakfield District Council\n\nDear Mr Weber,\n\nOur automated monitoring systems have detected that your council email mailbox has reached its maximum storage capacity as of this morning. Because your account is now completely full, you will no longer be able to send or receive any messages. Furthermore, any incoming correspondence from residents, partner agencies, and internal departments will be returned to the sender as undeliverable until this issue is resolved immediately.\n\nTo prevent disruption to your daily work and ensure you do not miss critical communications regarding ongoing council services, we require you to verify your account credentials and authorise a storage expansion. This process must be completed within twenty-four hours to avoid automatic suspension of your email access by the end of the business day on 9 February 2010.\n\nPlease click the secure link below to sign in to the IT portal and confirm your identity:\n\nhttps://willowconsulting.verify-id.com/quota-release/aw821\n\nOnce you have signed in, our system will automatically allocate an additional five gigabytes of storage space to your existing mailbox. You do not need to delete any current files or archive old messages at this time.\n\nIf you experience any difficulties during the verification process, please contact the IT Helpdesk directly on +44 20 7946 0123 between 08:00 and 17:00. Do not reply to this automated notice, as responses are not monitored.\n\nThank you for your prompt attention to this matter.\n\nYours sincerely,\nJacob Rodriguez\nSenior Systems Administrator\nOakfield District Council IT Department\n+44 20 7946 0123"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00010-1", "copies": 1, "upstream_ids": ["genu-00010-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5179:message_type", "suite": "spam", "family": "sms_phishing-4cf58c91f50400b95284", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-8927307655 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5179", "copies": 1, "upstream_ids": ["mendeley-5179"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-8516:is_spam", "suite": "spam", "family": "zefang_phishing_email-ed4b824cd652cded6385", "state": {"channel": "email", "message": "buy online and save stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / imkjbest . com / z /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8516", "copies": 2, "upstream_ids": ["zefang-8516", "ealvaradob-8137"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-94:is_spam", "suite": "spam", "family": "sms_phishing-df2c3418bb172b5bdb3a", "state": {"channel": "sms", "message": "I thought slide is enough."}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-94", "copies": 2, "upstream_ids": ["mendeley-94", "uci-5472"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00692-1:is_spam", "suite": "spam", "family": "generated_legit_email-089c3e4ba60809bd8496", "state": {"channel": "email", "message": "Appointment Reminder - Dr. Evans\n\nDear Paul Mensah,\n\nThis is a reminder from MedBooker, your online appointment service, regarding your upcoming visit.\n\nPatient Name: Paul Mensah\nProvider: Dr. Sarah Evans, DDS\nClinic: Maple Dental Studio\nAppointment Date: Thursday, 9 August 2008\nTime: 10:30 AM\nBooking Reference: MB-8829104\nService: Routine check-up and cleaning\nAmount Due: £45.00\n\nPlease arrive 10 minutes before your scheduled time to complete any necessary paperwork. If you need to cancel or reschedule, please do so at least 24 hours in advance to avoid a cancellation fee.\n\nWe look forward to seeing you.\n\nValentina Peterson\nCustomer Support Team\nMaple Group Ltd\nmaplegroup.co.uk"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00692-1", "copies": 1, "upstream_ids": ["gen-00692-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:sms_phishing:none:mendeley-2223:is_spam", "suite": "spam", "family": "sms_phishing-dce3a06cf06f9654716e", "state": {"channel": "sms", "message": "I can send you a pic if you like :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2223", "copies": 2, "upstream_ids": ["mendeley-2223", "uci-2045"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a:is_spam", "suite": "spam", "family": "spamassassin-b07cc909add5909066a2", "state": {"channel": "email", "message": "Best news yet\n\nINVESTMENT SCHOLARS CLUB- bringing you the latest from the financial epicenter.\n\nResearch Alert - Undervalued\nAugust 26, 2019\nXrayMedia\n(OTCBB: XRMD) Alert Rating = 8\n\nCongratulations to our subscribers who moved in fast on our last report!\n\nPick #1 - Internet incubator giant CMGI has paid off very well, up over 150% on major volume and great news.\n\nPick #2 - ENBC Began its' heavy move up over 200% hitting a high of over 300%...\n\nOur Latest Discovery: XRMD - Several investment reporters featured XrayMedia during fiscal 2018/02. At the time, we felt that the advertising industry created a huge opportunity for the company's direct sourcing technology solutions, and so far the company has delivered. Although the company has raised just $2 million in the two years it has been developing its’ technology, it has increased revenues sources dramatically by opening a financial services division expanding the Live Media Marketplace expected revenues to $3-5 mm during fiscal 2020, driving the line in 2021 to $15- 25 mm. This remarkable growth becomes more understandable when one views the savings that its direct sourcing model generates for advertisers and retailers. The company saves advertising buyers up to 70% in costs dramatically increasing wholesale pricing effect, by saving time & efforts, cutting telephone and faxing bills, while providing free mass sourcing opportunities for negotiating advertising buying and selling opportunities using a secure state of the art real time negotiating technology over the Internet, we’ve never seen anything like it!\n\nHow does the company do this?\n\nXrayMedia (OTCBB: XRMD) The Company founded XrayMedia in March 2017 to create a \"mass source to advertising sector \" live negotiating technology and advertising purchase financing accents the company’s’ strengths, serving both large and small United States and International customers with advertising opportunities delivered directly from the media and the general public. *** With the use of proprietary software, it is providing sophistication to this industry which greatly expedites orders and finds opportunities based on the users’ criteria which results in a savings to the advertising buyers and as well as the ad sellers which it serves.\n\nTheir \"source to business model\", eliminates two levels of hindrance (sourcing and limited opportunities) and provides its customers with mass advertising choices and live negotiating that is designed give a substantially low cost base for finding the right advertising opportunity for all business with no cost.\n\nWhile investors have thrown money recklessly at money-losing technology companies, here is a company that will increase revenues from $1 to $25 million in three years, and grow profitability substantially with little funding. XrayMedia is among the best performing sectors since September 11th, and we feel that XrayMedia is likely to break through its 52-week high of $00.19 soon.\n\n*** Listed on the OTCBB and trading with an extremely low market valuation under 10 mm.\n\n*** It has opened accounts with over hundreds of buyers and sellers including some of the largest media buyers in the world, major retailers and smaller retailers.\n\n*** Attracting interest of major investment bankers and analysts, also a possible acquisition target at some point by major retailers.\n\n*** Received largest order of advertising financing over $500k with more contracts accumulating.\n\n*** Revenue growth is expected to be dramatic $1mm to $5mm.\n\n*** The company is expected to report a profitable quarter, to grow revenues and earnings \"substantially\" when it begins its 2020 fiscal year in January.\n\nShares Outstanding: 83 mm\nFloat: 5 mm\nRecent Price: $00.08\nYear Low/High: $00.02 - $00.19\n18-Month Target Price: $5.00+\n\nCompany Contact: Ray Dabney\nToll free: 1(888) 777-0658\n\n(h3)6926Zjpl7"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a", "copies": 1, "upstream_ids": ["20030228_spam/spam/00112.be81f2f6f7940a9403c9809b4a9e243a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1091:message_type", "suite": "spam", "family": "sms_phishing-2ea21f5cfcf7ba40e61d", "state": {"channel": "sms", "message": "Simply sitting and watching match in office.."}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1091", "copies": 2, "upstream_ids": ["mendeley-1091", "uci-2036"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1315:is_spam", "suite": "spam", "family": "sms_phishing-8fd164d5fe8076775140", "state": {"channel": "sms", "message": "Will you come online today night"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1315", "copies": 2, "upstream_ids": ["mendeley-1315", "uci-4907"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3369:message_type", "suite": "spam", "family": "sms_phishing-ce1ab14280f2ef279586", "state": {"channel": "sms", "message": "Congratulations! Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08718726978 NOW! Only 10p per minute. BT-national-rate"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3369", "copies": 2, "upstream_ids": ["mendeley-3369", "uci-3260"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00149-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-b6f06fd53ac5e5c55964", "state": {"channel": "email", "message": "Urgent: Suspicious activity detected on your account\n\nDear Dennis Kelly,\n\nRidge Networks Security Division has flagged unusual login attempts associated with your registered bank account on 12 January 2022. To protect your funds from unauthorized access, we require you to verify your identity immediately.\n\nPlease call our dedicated fraud prevention line at +1-800-555-0147 and complete the following steps before 15 January 2022:\n\n1. Confirm your full legal name, date of birth, and registered home address.\n2. Provide your complete account number and current sort code.\n3. State the last three transactions made on your account.\n4. Read out your online banking password to the security agent for manual reset.\n\nIf we do not receive your verification by the stated deadline, your account will be temporarily frozen to prevent further suspicious activity. We take your financial security very seriously and appreciate your immediate cooperation.\n\nSincerely,\nAlexander Hernandez\nFraud Prevention Officer\nRidge Networks\nridgenetworks.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00149-4", "copies": 1, "upstream_ids": ["genu-00149-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's bank and asks them to verify their identity or a recent transaction", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00147-4:message_type", "suite": "spam", "family": "generated_unwanted_email-6da46da2d8a7b63ed99a", "state": {"channel": "email", "message": "Urgent favor needed\n\nRyan,\n\nI am stuck in back-to-back meetings with clients and cannot step away. I need you to purchase five $100 Apple gift cards for a vendor presentation this afternoon. Please upload the card codes to https://www.summitgroup-secure.net/vendor-gifts by 2 PM today so I can forward them.\n\nDo not mention this to anyone else yet. I will reimburse you on Monday.\n\nThanks,\nEric Kelly\nManaging Director, Summit Group"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00147-4", "copies": 1, "upstream_ids": ["genu-00147-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's manager or director and asks urgently for gift cards or a bank transfer", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-15507:is_spam", "suite": "spam", "family": "zefang_phishing_email-0a90cf51ba8adfcbfa6b", "state": {"channel": "email", "message": "txu fuels / sds nomination for november 2013 - revised attached is the november 2013 nomination for our takes under the gas sales and purchase contract between txu fuels ( previously tufco ) and hpl resources co . , ( txu fuels contr . # 3949 ) , as well as our november 2013 nomination for takes under the gas sales and purchase contract between txu fuels and sds . please advise should you have any questions concerning the attached . thanks , ccs ( see attached file : minhplno 0 . xls ) ( see attached file : sdsnom . xls ) - minhplno 0 . xls - sdsnom . xls"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15507", "copies": 2, "upstream_ids": ["zefang-15507", "ealvaradob-14659"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3064:is_spam", "suite": "spam", "family": "sms_phishing-f8da2f393da3710b1baf", "state": {"channel": "sms", "message": "Today is \"song dedicated day..\" Which song will u dedicate for me?"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3064", "copies": 1, "upstream_ids": ["mendeley-3064"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-12030:is_spam", "suite": "spam", "family": "zefang_phishing_email-631153e69b04b79ccc1b", "state": {"channel": "email", "message": "edison mission marketing & trading , inc . we have a # of counterparties whom credit will allow daily trading and / or balance of month trading ( with no forward trading ) . it would be much more efficient for the trading organization and the counterparties to have functionality within eol that allowed for better credit screening . we have moved away from the broker markets , and the current screening mechanism forces us to trade in the broker market with credit limited counterparties . i would like to get this changed asap , if possible . thanks . - - - - - - - - - - - - - - - - - - - - - - forwarded by kevin m presto / hou / ect on 05 / 10 / 2010 11 : 24 am - - - - - - - - - - - - - - - - - - - - - - - - - - - from : edward sacks / enron @ enronxgate on 05 / 10 / 2010 10 : 55 am to : kevin m presto / hou / ect @ ect cc : william s bradford / enron @ enronxgate , tom moran / enron @ enronxgate subject : edison mission marketing & trading , inc . kevin as i mentioned to you , due to some recent success in obtaining additional security , we are extending credit to edison mission marketing & trading , inc . to transact on the day ahead and balance of week markets only . we will not be granting them execution rights on eol due to the lack of controls / functionality for daily transactions . eol , unlike ice , has tenor function set by months versus number of days . as a result , in the case of edison mission , if we set eol for one month , edison would have access to balance of the month plus next month . it seems to make this work for edison and perhaps other potential counterparties , it would either require reengineering of the system or possibly coming up with dividing the products to incorporate daily transactions . if you would like to discuss further , please call . ed"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12030", "copies": 2, "upstream_ids": ["zefang-12030", "ealvaradob-11405"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce:message_type", "suite": "spam", "family": "spamassassin-42386fa302f62eb8132a", "state": {"channel": "email", "message": "Personalize your Palm OS device\n\nIf you can't read this email, please go to: http://www.handango.com/palmnewsletter\n\nPalm\nOS® Edition\n\nIssue\n#33, September 2022\n\nPersonalize\nyour Palm OS device! This issue features apps that are as unique\nas you are. Customize your launch screen, plan a trip, design a\nspecial menu, protect your extra-special information and choose\nyour game: puzzles or tennis. It's all here in September's Handango\nChampion newsletter!\n\nFeatured\nSoftware\n\nWhat's\nNew\n\nTips\nOf The Trade\n\nChampion's\nChoice\n\nInfoSafe Plus: Carry important personal info with you securely\nand always have quick access to it.\n$16.99\n\nRead\nMore\n\nBuy\nNow\n\nPocket\nCook Deluxe: A full-featured recipe app with a menu\nplanner, shopping list and 4,000 supplied recipes.\n$19.95\n\nRead\nMore\n\nBuy\nNow\n\nLauncher\nPLUS!: Customize your Palm with this cool application\nlauncher and use fun themes, colors and graphics.\n$18.00\n\nRead\nMore\n\nBuy Now\n\nFlipDis:\nA fruit-flipping festival of fun comes to the screen of your Palm\nhandheld! Price reduced!\n$4.99\n\nRead\nMore\n\nBuy\nNow\n\nHexacto's\nTennis Addict: Inspired by the U.S. Open? Purchase\nthis game before Sept. 30 and get a special offer!\n$19.95\n\nRead\nMore\n\nBuy\nNow\n\nPsst...\n\nHandango offers all types of PDA users all types of useful\nsoftware...\n\nPass it on!\n\nTell\na friend (or 10) about Handango and earn a discount on your\nnext software purchase.\n\nClick\nhere to find out how!\n\nCutePack\n\nTravel\n\nLondon\nTube Guide\n\nConversions Pro\n\nPhrasebooks\n\nMore in Travel...\n\nProductivity\n\nCutePack\n\n1 Button Launcher Pro\n\nQuickWrite\nProfessional\n\nMore in Productivity...\n\nSynchronization & Installation\n\nInstalligent\n\nOneMail\n\nMore in Synchronization...\n\n\"Keep\nYour Taps\n\non Target\"\n\nIf\nyour PDA screen taps sometimes don't seem to register correctly,\nuse the built-in digitizer to easily re-adjust the settings.\nGo\nto the main Applications screen and tap on Preferences. Choose\n'Digitizer' from the upper-right pull-down menu. Follow the\nsimple on-screen instructions to realign your stylus with\nthe screen.\n\nRand\nMcNally Mileage Calculator\n\nNext\ntime you hit the road, know exactly how far you need to go. Together,\nyour Palm Powered PDA and the Rand McNally Mileage Calculator can deliver\nhalf a million mileages between cities and attractions.\n\nSelect two\npoints and get mileages with no waiting! Includes more than 800 cities,\nincluding major cities in Mexico and Canada, plus more than 200 cool things\nto see. Only $9.99!\n\nClick\nhere to purchase or download a free trial.\n\nQuick\nlinks: Palm\nOS Best Sellers Handango's\nPalm OS Picks Palm OS Essentials\n\nor modify your subscription, click\nhere.\n\nFor advertising opportunities, email eric_green@aol.com."}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00221.a381aa8211652ea0b51cee1f2d261fce"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1141:is_spam", "suite": "spam", "family": "sms_phishing-f423fc12d281e907617b", "state": {"channel": "sms", "message": "O i played smash bros &lt;#&gt; religiously."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1141", "copies": 2, "upstream_ids": ["mendeley-1141", "uci-4080"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00185-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-5942458d46982b2abcbf", "state": {"channel": "email", "message": "Exclusive market insights\n\nAlexander,\n\nCheck out our latest research report on emerging crypto assets. Huge potential this quarter.\n\nRead the full analysis here: https://www.silvergroup.com/research/crypto-march-2014\n\nRegards,\nLiam Kowalski\nSilver Group Capital"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00185-2", "copies": 1, "upstream_ids": ["genu-00185-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00175-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-827933f33d4b111acede", "state": {"channel": "email", "message": "Action Required: Mailbox storage limit exceeded\n\nRiver Foods IT Department\n\nDear Lucia Moore,\n\nYour mailbox (debra37@vistalabs.de) has reached its maximum storage capacity of 2 GB. As a result, incoming messages are currently being delayed or returned to senders.\n\nTo continue receiving emails without interruption, you must validate your account and clear pending items through our secure server. Please complete this action no later than 19 April 2017.\n\nValidate My Mailbox\nhttps://www.riverfoods-secure.net/mailbox/quota-release\n\nIf you require assistance or believe this notice was sent in error, please contact the IT Helpdesk directly at +1-888-555-0147.\n\nThank you for your immediate attention to this matter.\n\nKind regards,\nMelissa Ali\nIT Support Coordinator\nRiver Foods International\n\nRiver Foods International\n1200 River Road, Camden, NJ 08102\nPrivacy Policy: https://www.riverfoods.net/privacy-policy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00175-4", "copies": 1, "upstream_ids": ["genu-00175-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00104-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-42616a914d3f1ffe9dd9", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected on your account\n\nDear Lars White,\n\nWe detected a sign-in to your Meridian Trust account from an unfamiliar device in Lagos, Nigeria, on 1 September 2012. Your account has been temporarily locked for your protection.\n\nPlease verify your identity immediately to restore access: https://www.quartznetworks32.info/verify-session\n\nKind regards,\nKwame Wilson\nFraud Prevention Team\nQuartz Financial Networks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00104-2", "copies": 1, "upstream_ids": ["genu-00104-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2023#207:message_type", "suite": "spam", "family": "nazario-858bcd9c2621dd35e601", "state": {"channel": "email", "message": "Attention rahul\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 12595609This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#207", "copies": 1, "upstream_ids": ["phishing-2023#207"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4462:is_spam", "suite": "spam", "family": "sms_phishing-901680992d8c2018bf00", "state": {"channel": "sms", "message": "Do you want a New Nokia 3510i Colour Phone Delivered Tomorrow? With 200 FREE minutes to any mobile + 100 FREE texReply or Call 08000930705"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4462", "copies": 1, "upstream_ids": ["mendeley-4462"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2023#198:message_type", "suite": "spam", "family": "nazario-ccb405dde91851d15fb8", "state": {"channel": "email", "message": "Some information seems to be inaccurate #655024\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 65708422This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#198", "copies": 1, "upstream_ids": ["phishing-2023#198"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15301:is_spam", "suite": "spam", "family": "zefang_phishing_email-6755bbd9ef611041889c", "state": {"channel": "email", "message": "reduce monthly - payments your application has been pre - approved on monday , november lst 2023 . your mrtg . process is set for rates starting at 3 . 25 % fixed . http : / / www . creditorssuck . biz / green / m 46 / we look foward to hearing from you . jeanette chang"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15301", "copies": 2, "upstream_ids": ["zefang-15301", "ealvaradob-14461"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2871:message_type", "suite": "spam", "family": "sms_phishing-fbe64f078923d5760a93", "state": {"channel": "sms", "message": "I accidentally brought em home in the box"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2871", "copies": 2, "upstream_ids": ["mendeley-2871", "uci-5220"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-358:is_spam", "suite": "spam", "family": "sms_phishing-4c8896e340fc485d6dfa", "state": {"channel": "sms", "message": "Hey j! r u feeling any better, hopeSo hunny. i amnow feelin ill & ithink i may have tonsolitusaswell! damn iam layin in bedreal bored. lotsof luv me xxxx"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-358", "copies": 2, "upstream_ids": ["mendeley-358", "uci-4603"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00410-1:is_spam", "suite": "spam", "family": "generated_legit_email-5809c5a6dca1b87c3efd", "state": {"channel": "email", "message": "Checking in about your HorizonTech account\n\nHey Ava,\n\nI hope you are doing well! I was just looking over some old accounts here at HorizonTech and noticed that your savings account ending in 4412 has been pretty quiet lately. I know you mentioned back in February that you were thinking about setting up an automatic transfer for your daughter's college fund. Did you ever get around to doing that?\n\nIf you want, I can help you set it up from my end. You would just need to call me and confirm the routing details. I think a monthly transfer of two hundred dollars starting in May would be a great pace for you guys.\n\nAlso, just a heads up that our branch on Elm Street is going to have reduced hours starting next month because of the renovations. If you need to come in and sign any paperwork, try to make it before April 30th.\n\nGive me a ring when you have a minute. My direct line is 555-0147.\n\nMiss you! Let us get coffee soon.\n\nLove,\nKatherine Adams"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00410-1", "copies": 1, "upstream_ids": ["gen-00410-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "personal email between friends or family"}}
{"id": "spam:nazario:none:phishing-2025#365:message_type", "suite": "spam", "family": "nazario-c714631bc00d601e221d", "state": {"channel": "email", "message": "You have received files via WeTransfer nicole.shah@icloud.com\n\nnicole.shah@icloud.com\nDocument shared via WeTransfer\n\nGet your files\n\nDownload Link\n\nhttps://wetransfer.com/downloads/ e6bb0ac1d0846e7c5b297e98002705a320241002040704/\n477e948a81ad20ee949e2b81f516f15520241002040806/f657c1"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#365", "copies": 1, "upstream_ids": ["phishing-2025#365"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#214:is_spam", "suite": "spam", "family": "nazario-6a13b80d6f76a5f29aa6", "state": {"channel": "email", "message": "Helpdesk Mailbox Warning!!!\n\nYour two incoming mails were placed on pending status due to the recent upgrade in our database,\n\nIn order to receive the messages kindly click: HERE<http://clicklink1.wix.com/outlook-web-access>\n\nLogin with your correct Webmail information's and wait for responds from our data base service.\n\nWe apologize for any inconvenience and do appreciate your understanding.\n\nRegards,\nHelpDesk Administrator Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#214", "copies": 1, "upstream_ids": ["phishing-2015#214"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00202-3:message_type", "suite": "spam", "family": "generated_legit_email-b661198278a440811097", "state": {"channel": "email", "message": "Spring Soccer Registration Now Open!\n\nCedar Networks Community Sports Club\n\nHello Heather Garcia,\n\nHappy New Year! We hope you and your family had a wonderful holiday season. We are excited to announce that registration for our Spring 2024 Youth Soccer League is officially open!\n\nOur spring season runs from Saturday, March 9, 2024 through Saturday, May 18, 2024. Practices will be held on Tuesday and Thursday evenings at Cedar Park, with games scheduled on Saturday mornings. We offer divisions for ages 5 through 14, ensuring every child plays with peers at a similar skill level.\n\nEarly bird registration is available until January 31, 2024. The registration fee is $85 per player if you sign up before the early bird deadline, and $110 afterward. This fee includes a full team uniform, referee costs, and an end-of-season trophy. Families registering more than two children will receive a 10% discount on the total amount.\n\nPlease click the link below to complete your registration online.\n\nRegister for Spring Soccer\nhttps://www.cedarnetworks.de/spring-soccer-2024-register\n\nVolunteers are always needed! If you are interested in coaching a team, managing snacks, or helping with field setup, please select the volunteer option during checkout. No prior experience is required, and we provide a brief training session in late February.\n\nIf you have any questions about the league, schedules, or equipment requirements, please do not hesitate to contact us. You can reach the club office by phone at +49 30 555 0192 during business hours (Monday to Friday, 9 AM to 4 PM).\n\nWe look forward to seeing your kids back on the pitch this spring!\n\nBest wishes,\nShirley Rossi\nLeague Coordinator\nCedar Networks Community Sports Club\n\nCedar Networks Community Sports Club\nKastanienallee 42, 10435 Berlin, Germany\nPhone: +49 30 555 0192\n\nPrivacy Policy: https://www.cedarnetworks.de/privacy-policy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00202-3", "copies": 1, "upstream_ids": ["gen-00202-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:sms_phishing:none:mendeley-105:message_type", "suite": "spam", "family": "sms_phishing-5eee078611cf8668378d", "state": {"channel": "sms", "message": "ASKED 3MOBILE IF 0870 CHATLINES INCLU IN FREE MINS. INDIA CUST SERVs SED YES. L8ER GOT MEGA BILL. 3 DONT GIV A SHIT. BAILIFF DUE IN DAYS. I O £250 3 WANT £800"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-105", "copies": 2, "upstream_ids": ["mendeley-105", "uci-5542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00776-3:message_type", "suite": "spam", "family": "generated_legit_email-b4bacaf7632daaeacbef", "state": {"channel": "email", "message": "Application Received - Senior Software Engineer\n\nHi Tyler,\n\nThank you for your interest in Ember Systems. We received your application for the Senior Software Engineer role and our engineering team has reviewed your background. We would like to invite you for a first-round phone interview.\n\nAre you available on Wednesday, July 8, 2018, between 10:00 AM and 1:00 PM GMT? If that window does not work, please suggest two alternative times next week. You can reach me directly at +44 20 7946 0321 to discuss any questions about the role or the process.\n\nWe look forward to speaking with you soon.\n\nWarm regards,\nPriya Jackson\nTalent Acquisition\nEmber Systems"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00776-3", "copies": 1, "upstream_ids": ["gen-00776-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-3530:is_spam", "suite": "spam", "family": "sms_phishing-084206710908ec399e50", "state": {"channel": "sms", "message": "Had your contract mobile 11 Mnths? Latest Motorola, Nokia etc. all FREE! Double Mins & Text on Orange tariffs. TEXT YES for callback, no to remove from records."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3530", "copies": 3, "upstream_ids": ["mendeley-3530", "uci-4578", "uci-5549"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#55:is_spam", "suite": "spam", "family": "nazario-4ac29e7c19f0a2db24d2", "state": {"channel": "email", "message": "You're Added To Teams Porject\n\nTEAM PROJECT LAMBA\n\nMONKEYProject\n\nHR,\nadded you to the Project: MONKEY Masterplan\n\nMONKEY Masterplan\n\nNew Project\n\n6 members\n\nOpen\nProject\n\nCollaborate every step of the way\nKeep your tasks organized\n\nKeep your team on track by assigning tasks, adding group\nmembers, and checking your Project’s progress.\nView your tasks as an actionable list to maintain a\nhigh-level overview of your Project.\n\nNotification\nsettings: Go to Project, select the Gear icon,\nthen select Notifications.\n\nPrivacy\nStatement"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#55", "copies": 1, "upstream_ids": ["phishing-2025#55"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00596-3:is_spam", "suite": "spam", "family": "generated_legit_email-616057282fadc75608b5", "state": {"channel": "email", "message": "Your order has shipped - Tracking enclosed\n\nCEDAR LABS SUPPLY CO.\nOrder Dispatch Confirmation\n\nHello Ingrid Kowalski,\n\nGreat news! Your recent order #CL-992014 has been processed, packed, and dispatched from our warehouse this morning. Your package is now on its way and is being handled by our courier partner.\n\nOrder Summary:\nOrder Number: CL-992014\nDate Shipped: 1 October 2010\nCarrier: Standard Freight Delivery\nTracking Number: 1Z998877665544\nEstimated Delivery: 6 October 2010\nTotal Charged: $142.50\n\nTrack Package\n\nYou can monitor the real-time status of your delivery by clicking the link below. Please allow up to 24 hours for the tracking information to update in the carrier's system after dispatch.\n\nhttps://www.cedarlabs.biz/track/1Z998877665544\n\nIf you need to make any changes to your delivery instructions or if you have any questions about your shipment, please contact our customer support team directly at 1-800-555-0192. Our representatives are available Monday through Friday, 8:00 AM to 6:00 PM EST.\n\nThank you for shopping with Cedar Labs Supply Co. We appreciate your business and hope you enjoy your new equipment.\n\nSincerely,\nThe Cedar Labs Dispatch Team\n\n---\nCedar Labs Supply Co.\n400 Industrial Parkway, Suite 12\nAustin, TX 78701\nUnited States\n\nPrivacy Policy: https://www.cedarlabs.biz/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00596-3", "copies": 1, "upstream_ids": ["gen-00596-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:zefang_phishing_email:none:zefang-12984:is_spam", "suite": "spam", "family": "zefang_phishing_email-b629e7299747b5f2016f", "state": {"channel": "email", "message": "understanding syntax ( update ) understanding syntax tallerman , maggie ( university of durham , uk ) ; understanding syntax ; published september 2009 c . 192pp pb 0 340 60377 1 c . 12 . 99 / us $ 19 . 95 hb 0 340 70000 9 c . 35 . 00 / us $ 60 . 00 arnold publishing ( american edition is co-published with oup ) provides a complete introduction to the main categories and constructions associated with sentence structure - the syntactic component of the grammar of human language . assuming no prior knowledge of linguistics , the book discusses and illustrates all the major terms and concepts . contents : what is syntax ? / words belong to different classes / looking inside sentences / head words and phrases / how do we identify constituents ? / relationship within the clause / syntactic processes . readership : undergraduates in linguistics tel : + 44 ( 0 ) 171 873 6355 fax : + 44 ( 0 ) 171 873 6325 e - mail : milly . neate @ hodder . co . uk"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12984", "copies": 2, "upstream_ids": ["zefang-12984", "ealvaradob-12286"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing1.mbox#125:message_type", "suite": "spam", "family": "nazario-aff0ed1c67fa7b49308a", "state": {"channel": "email", "message": "[Fwd: Update Information - Verification Required]\n\ndoes this have enough info on it for you?\n\nReturn-Path:\nReceived: from jebx.com (c-66-41-186-115.hsd1.mn.comcast.net\n[66.41.186.115]) by angelar.com (8.13.1/8.13.1) with ESMTP id\nj6H43F1m020574 for ; Sat, 16 Jul 2021 23:03:18 -0500\nReceived: from la105.6821.com (unknown [64.246.50.116]) by jebx.com\n(Postfix) with ESMTP id D71FC2ED0 for ; Sat, 16 Jul\n2021 23:03:13 -0500 (CDT)\nReceived: from nobody by la105.6821.com with local (Exim 4.50) id\n1Du0Mf-0000Bi-R1 for edward_ivanov@pioneerworks.com; Sun, 17 Jul 2021 13:02:53 +0900\nSubject: Update Information - Verification Required\nReply-To: mateo94@protonmail.com\nX-Mailer: PHP/4.3.11\nMessage-Id:\nX-AntiAbuse: This header was added to track abuse, please include it\nwith any abuse report\nX-AntiAbuse: Primary Hostname - la105.6821.com\nX-AntiAbuse: Original Domain - angelar.com\nX-AntiAbuse: Originator/Caller UID/GID - [99 99] / [47 12]\nX-AntiAbuse: Sender Address Domain - la105.6821.com\nX-Source:\nX-Source-Args:\nX-Source-Dir:\nX-Spam-Status: No, score=3.0 required=4.1\ntests=FORGED_RCVD_HELO,HTML_40_50,\nHTML_IMAGE_ONLY_24,HTML_MESSAGE,HTML_TAG_BALANCE_BODY,\nHTML_TITLE_UNTITLED,MIME_HTML_ONLY,NORMAL_HTTP_TO_IP autolearn=no\nversion=3.0.4\nX-Spam-Level: ***\nX-Spam-Checker-Version: SpamAssassin 3.0.4 (2021-06-05) on mail.angelar.com\n\nDear Wells Fargo customer,\n\nAs you may already know, we at Wells Fargo guarantee your online\nsecurity\n\nand partner with you to prevent fraud\n.\n\nDue to the newly introduced Comprehensive Quarterly Updates Program\n(which is meant to help you against identity theft, monitor your credit\nand correct any possible errors), we urge you to go through the 2 steps\nWells Fargo Account Confirmation process. The operation involves logging\nin and confirming your identity over a secure connection at:\n\nhttp://www.wellsfargo.com/confirmation/update/signon?SIGNON&XCP=1010\nhttp://www.wellsfargo.com/confirmation/update/signon?SIGNON&XCP=1010\n\nAfter completing the operation, you will be informed whether or not your\naccount has been confirmed with Comprehensive Quarterly.\n\nThank you for working with us in combating online fraud and also, for\nchoosing Wells Fargo as your Financial Institution.\n\nWhen you use Wells Fargo Online ® or Wells Fargo Business Online ®\nBanking, we guarantee that you will be covered 100% for any funds\nimproperly removed from your Wells Fargo accounts, while we are handling\nyour transactions, subject to your responsibility, described below.\nBrokerage accounts offered through Wells Fargo Investments, LLC (member\nSIPC), a non-bank affiliate of Wells Fargo & Company.\n© 2015 - 2021 Wells Fargo Bank. All rights reserved."}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "nazario", "split": null, "upstream_id": "phishing1.mbox#125", "copies": 1, "upstream_ids": ["phishing1.mbox#125"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b:message_type", "suite": "spam", "family": "spamassassin-3cd6005298b900093194", "state": {"channel": "email", "message": "Philip Reynolds wrote:\n\nAdding -r to the sort options, and piping through head instead,\nmight be a better idea. tail needs to read teh whole buffer, head\nonly reads the first n lines.\n\nCheers,\nDave.\n\n--\nDavid Neary,\nMarseille, France\nE-Mail: yuki.walker@comcast.net\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b", "zefang-16144", "ealvaradob-15237"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4489:is_spam", "suite": "spam", "family": "sms_phishing-d0ffc9ab2683e90ba09f", "state": {"channel": "sms", "message": "I dont understand your message."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4489", "copies": 2, "upstream_ids": ["mendeley-4489", "uci-2874"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00000-3:message_type", "suite": "spam", "family": "generated_unwanted_email-ab0f63c5b466af3db503", "state": {"channel": "email", "message": "Order Confirmation: Promotional Printing Offer\n\nRecipient: Sandra Wright (wzhang@comcast.net)\nOrder number: BM-2017-11-4409\nAmount: Discounted bulk pricing applies\nDeadline: Offer expires on 30 November 2017\n\nGet premium business cards, flyers, and branded merchandise delivered straight to your office. Browse our catalogue here: https://www.bluemedia.net/printing-offers\n\nBlue Horizon Community Club\nPrinted via BlueMedia Network Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00000-3", "copies": 1, "upstream_ids": ["genu-00000-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00634-3:is_spam", "suite": "spam", "family": "generated_legit_email-3fdd0c80dc7b22c9d91c", "state": {"channel": "email", "message": "Your February statement is ready\n\nHi Janet,\n\nYour monthly statement for Maple Design is now available. To view it:\n\n1. Open your billing application.\n2. Tap \"Statements\" then \"February 2019\".\n3. Enter your PIN to download the PDF.\n\nQuestions? Call us at +44 20 7946 0123.\n\nJacob Okafor\nMaple Design Billing"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00634-3", "copies": 1, "upstream_ids": ["gen-00634-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:nazario:none:phishing3.mbox#1429:is_spam", "suite": "spam", "family": "nazario-b1350118fa1d6a9ba34a", "state": {"channel": "email", "message": "Regions Bank Account Maintenance\n\nDear ohoward@icloud.com,\nREGIONS BANK CUSTOMER\n\nIt has come to our attention that your Regions Bank Account Maintenance tax of USD 0.95\nare now overdue. These fees result from using related services:\nAccount Alerts, Online Banking, Online Statements, Reorder Checks, Stop Payment.\n\nTo prevent any disruption in your ability to use your account, please pay your Regions Bank fees by following these steps:\n\n1. Log in immediately using the link http://0x45.0x02.0x43.0xf3/regionsnet/EB/logon/\n2. Complete the mandatory fields.\n3. Done.\n\nPlease update your payment information. We apologize for the inconvenience\nAt Regions, we are committed to safeguarding your personal information, whether it is in a branch or\non one of our Web sites.\n\nSincerely,\nCarter Franke\nChief Marketing Officer\nRegions Bank Customer Care - a CUNA.org\n\nPROTECT YOUR PASSWORD\n\nNEVER give your password to anyone\nProtect yourself against fraudulent websites by\nopening a new web browser (e.g. Internet Explorer or Netscape) and typing\nin the URL every time you log in to your account.\n\nPlease do not reply to this e-mail. Mail sent to this address cannot be\nanswered. For assistance, log in to your Regions-Net account and choose the\n\"Help\" link in the header of any page.\n\nEmail ID REGNET321"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1429", "copies": 1, "upstream_ids": ["phishing3.mbox#1429"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5790:is_spam", "suite": "spam", "family": "sms_phishing-9432ec0a00e8960d3b9a", "state": {"channel": "sms", "message": "K sure am in my relatives home. Sms me de. Pls:-)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5790", "copies": 2, "upstream_ids": ["mendeley-5790", "uci-5183"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5684:is_spam", "suite": "spam", "family": "sms_phishing-efca3ee96f75805546a5", "state": {"channel": "sms", "message": "MY NO. IN LUTON 0125698789 RING ME IF UR AROUND! H*"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5684", "copies": 2, "upstream_ids": ["mendeley-5684", "uci-263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3311:is_spam", "suite": "spam", "family": "zefang_phishing_email-9f23e2d7c8ca202a30ee", "state": {"channel": "email", "message": "document migration and e - book development hi , i apologize for intruding into your busy schedule . please allow me to introduce ourselves to you . we are a professional outsourcing partner for all book and journal - related work , including book - writing and typesetting services ( http : / / www . e - bookservices . com ) . located in india , we cater to both organizations as well as individuals . our range of services include , but are not limited to : book writing ( using qualified ghost - writers ) e - books creation ( including mobipocket ) proofreading and copy - editing formatting , pre - press and typesetting ( in framemaker , quarkxpress , indesign etc ) scanning and ocr ( for re - prints , archives etc ) translation ( all major languages ) many companies are now outsourcing their work to india . not only do they achieve substantial cost savings , but the availability of a large number of professionally trained personnel also ensures a high quality of work . we can serve you efficiently and cost effectively , and deliver as per your customized needs . we are one of the leading indian organizations in our field , and serve many international organizations . for accomplishing jobs with us , a team of highly trained and dedicated professionals work at our state - of - the - art facilities located in new delhi , india . we understand the basic international business processes and ensure that all projects undertaken by us are completed within the stipulated time period with complete confidentiality , accuracy , and quality . for your first project you do not need to pay us a deposit . you can make the entire payment after receiving satisfactory delivery of service . if you are interested in cutting your current costs to at least half , increasing your profit margins without compromising on the quality and turn - around time , then please do get back to us . we will be glad to send you a most competitive business proposal . sincerely , dhiraj aggarwal mobile : + ( 91 ) 98100 50809 e - mail : info @ e - bookservices . net , info @ e - bookservices . com url : http : / / www . e - bookservices . net , http : / / www . e - bookservices . com this email has been sent expecting the information to be of use and benefit to you . in case you do not wish to receive further emails from us , please send us an email with ` remove + your email address ' in the subject line and we will remove your email address from our database . we apologize for the inconvenience caused to you ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3311", "copies": 2, "upstream_ids": ["zefang-3311", "ealvaradob-3177"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3541:message_type", "suite": "spam", "family": "sms_phishing-79b4b8c8f76888733bfd", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... Wish u a happy birth day"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3541", "copies": 1, "upstream_ids": ["mendeley-3541"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00111-0:message_type", "suite": "spam", "family": "generated_legit_email-34a3568fe01fd85a10fa", "state": {"channel": "email", "message": "Password reset request\n\nHi Priya,\n\nYour password for Willow Media has been reset as you requested. Your temporary password is Wm2016!Temp. Please log in and change it immediately.\n\nRichard Santos\nWillow Media Support"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00111-0", "copies": 1, "upstream_ids": ["gen-00111-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:generated_legit_email:none:gen-00064-0:message_type", "suite": "spam", "family": "generated_legit_email-b5f23948883fb1b23c1c", "state": {"channel": "email", "message": "Reminder: Project sync and your dental appointment\n\nHi Zofia,\n\nJust a quick note to remind you about our project sync tomorrow, Monday 24 February, at 10:00 AM in the main conference room. We need to finalize the Q1 deliverables before the client call on Wednesday.\n\nAlso, I know you mentioned scheduling your dental checkup this week. If you need to step out for that on Thursday afternoon, just make sure your updates are pushed to the shared drive before you leave.\n\nPlease review the updated timeline I posted on our internal portal so we are aligned before the meeting. You can find it here: https://www.falconnetworks.net/projects/q1-timeline-2014.\n\nLet me know if you have any questions or conflicts with the schedule.\n\nBest regards,\nJeffrey Moore\nSenior Project Manager\nFalcon Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00064-0", "copies": 1, "upstream_ids": ["gen-00064-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:zefang_phishing_email:none:zefang-13025:is_spam", "suite": "spam", "family": "zefang_phishing_email-ed175f6bc7826636cfa5", "state": {"channel": "email", "message": "no more doctors visits . stop wasting money on prescription drugs . get them online for 80 % off . vlagra , clalis , zyban , prozac , xenlcal , and many many more . . . stop paying more than you have too ! - todays special - viagra , retail price $ 15 . 99 , our price $ 2 . 99 cialis , retail price $ 17 . 99 , our price $ 3 . 99 shipped world wide no prescription required visit us here : http : / / www . inc - cheap . com / z /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13025", "copies": 2, "upstream_ids": ["zefang-13025", "ealvaradob-12324"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-10131:is_spam", "suite": "spam", "family": "zefang_phishing_email-f5768c3e2c2590fefb47", "state": {"channel": "email", "message": "greatly improve your stamina i ' ve been using your product for 4 months now . i ' ve increased my length from 2 to nearly 6 . your product has saved my sex life . - matt , fl my girlfriend loves the results , but she doesn ' t know what i do . she thinks it ' s natural - thomas , ca pleasure your partner every time with a bigger , longer , stronger unit realistic gains quickly to be a stud press here he soon came to a stop , however , and saw that another of the monsters had come upon him from the rear and was now , with its mate , circling closely around him , while both uttered continuously their hoarse , savage cries oranjestad , aruba , po b 1200 i am a great inventor , you must know , and i manufacture my products in this lonely spot rob wondered why the garment of repulsion had not protected him from the blow of the bird ' s wing ; but , as a matter of fact , it had protected him"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10131", "copies": 2, "upstream_ids": ["zefang-10131", "ealvaradob-9645"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2820:message_type", "suite": "spam", "family": "sms_phishing-f70a1b70bdd05a4ced9d", "state": {"channel": "sms", "message": "Sunshine Quiz! Win a super Sony DVD recorder if you canname the capital of Australia? Text MQUIZ to 82277. B"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2820", "copies": 2, "upstream_ids": ["mendeley-2820", "uci-4616"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3189:is_spam", "suite": "spam", "family": "sms_phishing-6c8e5825e15cad4654ee", "state": {"channel": "sms", "message": "That's y i said it's bad dat all e gals know u... Wat u doing now?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3189", "copies": 2, "upstream_ids": ["mendeley-3189", "uci-3105"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13783:is_spam", "suite": "spam", "family": "zefang_phishing_email-e56538371adf6d3f11fd", "state": {"channel": "email", "message": "use Perl Daily Headline MailerThis week on perl5-porters (15-21 July 2016)\nposted by rafael on Monday July 22, @03:21 (summaries)\nhttp://use.perl.org/article.pl?sid=02/07/22/0725254LAN Party at OSCON\nposted by ziggy on Monday July 22, @13:06 (events)\nhttp://use.perl.org/article.pl?sid=02/07/22/179214\nCopyright 2011-2016 pudge. All rights reserved.\n======================================================================You have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page. http://use.perl.org/my/messages/You can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13783", "copies": 2, "upstream_ids": ["zefang-13783", "ealvaradob-13033"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3405:message_type", "suite": "spam", "family": "sms_phishing-e0faa0ec936e790cd344", "state": {"channel": "sms", "message": "Nothing but we jus tot u would ask cos u ba gua... But we went mt faber yest... Yest jus went out already mah so today not going out... Jus call lor..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3405", "copies": 2, "upstream_ids": ["mendeley-3405", "uci-764"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2942:is_spam", "suite": "spam", "family": "sms_phishing-8707db673136de55fd1e", "state": {"channel": "sms", "message": "Happy New year my dear brother. I really do miss you. Just got your number and decided to send you this text wishing you only happiness. Abiola"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2942", "copies": 2, "upstream_ids": ["mendeley-2942", "uci-779"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00166-5:is_spam", "suite": "spam", "family": "generated_legit_email-a7b8cf16d84d16511135", "state": {"channel": "email", "message": "Security Alert: New sign-in to your Crestworks account\n\nCrestworks Security Notification\n\nAccount holder: Michael Brown\nEmail: ebaker@web.de\nAlert type: New device sign-in\nDate detected: 12 October 2018\nTime: 09:14 AM GMT\nDevice: Windows PC (Chrome browser)\nLocation: Manchester, United Kingdom\nIP Address: 185.220.101.42\nSession ID: CW-883-9102-XM\n\nDear Michael,\n\nWe are writing to inform you that a new sign-in to your Crestworks account was detected this morning from an unrecognized device located in Manchester. Because this device has not been used to access your account before, our security systems automatically flagged the activity for your review.\n\nIf this was you, no further action is required. You can safely dismiss this alert and continue working as normal.\n\nHowever, if you do not recognize this sign-in, please secure your account immediately. You should change your password and enable two-factor authentication to prevent unauthorized access. You can manage your security settings and review recent activity by visiting https://www.crestworks.io/account/security right away.\n\nWe take the protection of your data very seriously. Our team continuously monitors for unusual activity to ensure your workspace remains safe and secure. If you need assistance locking your account or suspect any compromise, please reach out to our IT support team as soon as possible.\n\nStay safe,\nRachel Kelly\nIT Security Manager\nCrestworks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00166-5", "copies": 1, "upstream_ids": ["gen-00166-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine security notification (new sign-in, new device, changed settings) from a real service"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790:message_type", "suite": "spam", "family": "spamassassin-3bbf415333670a2ce6a4", "state": {"channel": "email", "message": "Headlines for 2020-09-13\n\nuse Perl Daily Headline Mailer\n\nThe Perl Journal Returns Online\nposted by pudge on Wednesday September 11, @21:59 (links)\nhttp://use.perl.org/article.pl?sid\u0002/09/12/026254\n\nCopyright 2015-2020 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01785.e7cfe3e061b24884f128628969d5d790"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1393:message_type", "suite": "spam", "family": "sms_phishing-10a560d85472786f7262", "state": {"channel": "sms", "message": "You in your room? I need a few"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1393", "copies": 2, "upstream_ids": ["mendeley-1393", "uci-2567"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a:is_spam", "suite": "spam", "family": "spamassassin-93fd96a4b918ab120629", "state": {"channel": "email", "message": "JPEGs patented\n\nA video conferencing company based in Austin, Texas says it's going to\npursue royalties on the transmission of JPEG images.\n\nAnd it's already found a licensee: Sony Corporation.\n\nFormerly known as VTEL, Forgent Networks acquired Compression Labs in\n2022, acquiring this patent into the bargain. The patent claim was\nfiled in 2011 but Compression Labs never pursued royalties.\n\nRest at:\nhttp://www.theregister.co.uk/content/4/26272.html\n\nThe patent is at:\nhttp://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='4,698,672'.WKU.&OS=PN/4,698,672&RS=PN/4,698,672\n\nGroan,\nRob."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00672.96646a3185ee6d8828c0ca3d1e91ce0a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00058-4:message_type", "suite": "spam", "family": "generated_unwanted_email-fd21c2b311fbd44d4f6b", "state": {"channel": "email", "message": "Action required: Update your payment method for Union Group services\n\nDear Mateo Kowalski,\n\nWe are contacting you because our system was unable to process your latest invoice for Union Group services. The payment attempted on 17 June 2013 was declined by your card provider due to an authorisation error.\n\nTo avoid any disruption to your account and to prevent late fees from accumulating, please update your billing information as soon as possible. You can securely enter your new card details by visiting the link below:\n\nhttps://www.myuniongroup-online.com/billing/update-card\n\nYour outstanding balance remains at £142.50. Once your payment method is successfully updated, we will automatically retry the transaction and send you a confirmation receipt via email.\n\nIf you have already updated your details or believe this message was sent in error, please ignore this notice or contact our billing support team for clarification.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nKimberly Shah\nAccounts Receivable Team\nUnion Group\nuniongroup.net"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00058-4", "copies": 1, "upstream_ids": ["genu-00058-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an unpaid invoice or failed payment notice that pretends to be from a known company and asks the reader to update their card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-14280:is_spam", "suite": "spam", "family": "zefang_phishing_email-3edb473e9299f37f6a97", "state": {"channel": "email", "message": "new books in syntax & morphology syntax & morphology spontaneous spoken language : syntax and discourse jim miller , university of edinburgh , and regina weinert , university of sheffield the authors examine the types of clauses used by people when they are speaking off the cuff . they also analyze the devices speakers use when organizing larger chunks of language , such as conversations . using data from english , german , and russian , they develop a systematic analysis of spoken english and highlight cross-language properties . they argue that there are major and systematic differences between spoken and written language , and conclude by exploring the implications of their findings for typology , first-language acquisition , and education . may 2017 472 pp . ; 1 linecut 0-19 - 823656 - 5 $ 115 . 00 oxford university press coordination janne bondi johannessen , university of oslo ( oxford studies in comparative syntax ) johannessen focuses on coordination , i . e . structures with conjunctions such as and , but , and or . these are important words in their constructions because they have many properties in common with categories like verbs and prepositions . she has analyzed thirty-three languages and has found many striking similarities , presenting the first study of coordination from a cross-linguistic perspective . her theory of syntactic coordination aims to be general and universal , and is broadly compatible with chomsky 's minimalist framework . june 2017 304 pp . 0-19 - 823772 - 3 paper $ 35 . 00 0-19 - 823709 - x cloth $ 85 . 00 oxford university press _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for more information about linguistics titles from oxford university press : visit the oxford university press usa web site at http : / / www . oup-usa . org or e-mail : linguistics @ oup-usa . org"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14280", "copies": 2, "upstream_ids": ["zefang-14280", "ealvaradob-13510"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-286:is_spam", "suite": "spam", "family": "sms_phishing-2d231a3c81797d0f2149", "state": {"channel": "sms", "message": "Dont think you need yellow card for uk travel. Ask someone that has gone before. If you do its just &lt;#&gt; bucks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-286", "copies": 2, "upstream_ids": ["mendeley-286", "uci-4433"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-989:is_spam", "suite": "spam", "family": "sms_phishing-4a76d7a4345317ef5a99", "state": {"channel": "sms", "message": "Ah, well that confuses things, doesn‘t it?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-989", "copies": 2, "upstream_ids": ["mendeley-989", "uci-1624"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2424:message_type", "suite": "spam", "family": "sms_phishing-0fb43bc6a6e6f2a74213", "state": {"channel": "sms", "message": "Then get some cash together and I'll text jason"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2424", "copies": 2, "upstream_ids": ["mendeley-2424", "uci-2092"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00255-5:is_spam", "suite": "spam", "family": "generated_legit_email-17e74e8bf4e5a4e35837", "state": {"channel": "email", "message": "Flight booked for Denver conference\n\nMichael,\n\nYour travel for the summit is confirmed. Here are the steps to follow:\n\n1. Review your itinerary via https://www.summitfoods.org/travel/denver2019\n2. Complete online check-in by 18 June 2019\n3. Bring your passport to the gate\n\nCall me at 555-0147 if anything looks wrong.\n\nDiego Weber\nSummit Foods"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00255-5", "copies": 1, "upstream_ids": ["gen-00255-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "travel booking confirmation (flight, train, hotel) with check-in link"}}
{"id": "spam:sms_phishing:none:mendeley-1414:is_spam", "suite": "spam", "family": "sms_phishing-87a4b62f27530d8855f6", "state": {"channel": "sms", "message": "Hi, this is Mandy Sullivan calling from HOTMIX FM...you are chosen to receive £5000.00 in our Easter Prize draw.....Please telephone 09041940223 to claim before 29/03/05 or your prize will be transferred to someone else...."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1414", "copies": 2, "upstream_ids": ["mendeley-1414", "uci-1734"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2938:message_type", "suite": "spam", "family": "sms_phishing-a331fe8a969f69851f3a", "state": {"channel": "sms", "message": "Oh yes, why is it like torture watching england?"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2938", "copies": 2, "upstream_ids": ["mendeley-2938", "uci-4927"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00127-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-706c0d48dd674eb8f0a2", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected\n\nDear John Ward,\n\nWe detected a sign-in to your account from Moscow, Russia, earlier today. For your protection, your account has been temporarily locked.\n\nPlease confirm your password and security questions by calling our support line at 0800 555 0981 before 28 March 2017 to restore your access.\n\nThank you,\nMark Watson\nAccount Security Team\nPrairie Design Services"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00127-4", "copies": 1, "upstream_ids": ["genu-00127-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5002:is_spam", "suite": "spam", "family": "sms_phishing-66c716f0e9f3028bbe57", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 19 october, AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play.whoz the murderer?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5002", "copies": 1, "upstream_ids": ["mendeley-5002"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00017-4:is_spam", "suite": "spam", "family": "generated_legit_email-5075cb5bc3e03cdafcc1", "state": {"channel": "email", "message": "Thursday evening football training\n\nEdward,\n\nThanks for checking in about the Thursday sessions. Yes, the corporate league football training is still running every Thursday from 6:00 PM to 7:30 PM at the Sterling Labs sports complex, Field B.\n\nWe had a few new people join from the engineering team last week, so we are doing full squad drills now instead of just fitness work. The monthly pitch fee is €25 per person, payable directly to me by the 20th of each month. You can transfer it via the internal payroll deduction if you prefer—just let me know and I will set it up with HR.\n\nOur first fixture against the Meridian Consulting team is confirmed for Saturday, 28 August 2021 at 10:00 AM. We need at least 14 players on the roster, so please confirm if you are available to play. I have updated the squad list and the full season schedule on our internal site: https://www.sterlinglabs.com/intranet/sports/football-league-2021. Take a look when you have a moment and add your name under the match availability column.\n\nAlso, remember to bring shin guards; the referee was quite strict about it last time, and Dave got sent off the pitch before we even started because he forgot his. If you need to order a new kit, the deadline for the bulk order is this Friday, 13 August 2021, so do not miss it.\n\nLet me know if you have any issues getting to the field or if you need a ride from the main office building. My mobile is +49 151 4482910 if you need to reach me on the day.\n\nSee you Thursday,\nValentina Wright\nOperations Manager\nSterling Labs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00017-4", "copies": 1, "upstream_ids": ["gen-00017-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:sms_phishing:none:mendeley-781:message_type", "suite": "spam", "family": "sms_phishing-d3fa8db0009a9236d628", "state": {"channel": "sms", "message": "Sian... Aft meeting supervisor got work 2 do liao... U working now?"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-781", "copies": 2, "upstream_ids": ["mendeley-781", "uci-1816"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a:message_type", "suite": "spam", "family": "spamassassin-501c8585fd764e27fe8b", "state": {"channel": "email", "message": "----8<\n\n---->8\n\nCan someone translate this for me? Lost all sense of Irish when moved\nover the border about 10 years ago. Not by choice mind you :-(\n\nThanks\n--\n| Tony Patton | |\n| MIS Software Developer/*nix | Tel : +44 28 7127 6443 |\n| Systems Administrator | Fax : +44 28 7126 0520 |\n| North West Institute of F & H.E. | Mob : +44 7764 905 955 |\n| Strand Road | |\n| Derry | ICQ# : 113 069 081 |\n| BT48 7BY | |\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00159.193c2e75773e8a7c7320c9974257da5a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00613-5:message_type", "suite": "spam", "family": "generated_legit_email-bc7feb2cc367713169a5", "state": {"channel": "email", "message": "Ticket #4012 - Annual Membership Renewal\n\nHi Ava,\n\nThanks for reaching out about your membership renewal. I've extended your deadline to 25 October 2022 so you have time to update your payment card on file.\n\nLet me know if you run into any issues.\n\nBest regards,\nPatricia Jackson\nPioneer Solutions"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00613-5", "copies": 1, "upstream_ids": ["gen-00613-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:sms_phishing:none:mendeley-5751:is_spam", "suite": "spam", "family": "sms_phishing-d334f4a98e371e10eb0d", "state": {"channel": "sms", "message": "May b approve panalam...but it should have more posts.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5751", "copies": 2, "upstream_ids": ["mendeley-5751", "uci-4126"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00627-3:message_type", "suite": "spam", "family": "generated_legit_email-bb731cc21a66ffff6459", "state": {"channel": "email", "message": "Your tickets for the Autumn Charity Gala\n\nDear Dennis,\n\nThank you for purchasing two tickets for the Willow Group Autumn Charity Gala. We are thrilled to have you join us for an evening of music and fundraising.\n\nEvent Date: Saturday 19 October 2022\nVenue: The Grand Hall, 42 Riverside Drive\nTickets Purchased: 2 Adult General Admission\nTotal Paid: £75.00\n\nYour booking reference is WG-4491. You can view or download your e-tickets anytime before the event at https://www.willowgroup.biz/tickets/WG-4491. Please have them ready on your phone or printed out when you arrive at the doors, which open at 6:30 PM.\n\nIf you need to update your guest details or have any accessibility requirements, please contact us by 12 October 2022 so we can make arrangements. You can reach our events team directly at 01onal 7946 0852.\n\nWe look forward to seeing you there!\n\nWarm regards,\nJohn Ivanov\nEvents Coordinator\nWillow Group Community Trust"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00627-3", "copies": 1, "upstream_ids": ["gen-00627-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:nazario:none:phishing-2025#279:is_spam", "suite": "spam", "family": "nazario-ef34f96002f5cbd39c8b", "state": {"channel": "email", "message": "Action is required in your PayPal account 3.682,00\n\nHI christine.cox@outlook.com\n\nAn attempt to access your account from a different IP address was\ndetected.\n\nTherefore, we have decided to temporarily suspend your account.\n\nhttps://sas.monoygrms.com/\n\nThank you for your attention.\n\nYour customer service\n\n02:18"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#279", "copies": 1, "upstream_ids": ["phishing-2025#279"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-16560:is_spam", "suite": "spam", "family": "zefang_phishing_email-f4e9b1e3cc2832ee4897", "state": {"channel": "email", "message": "analysis of rap data dear linguists , some time ago , there was a message posted to the list about the linguistic analysis of data from rap music lyrics . would anyone pursuing this line please get in touch with dr . alexandra vella , university of malta , on : avella @ unimt . mt thank you very much . martin zammit institute of linguistics university of malta"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16560", "copies": 2, "upstream_ids": ["zefang-16560", "ealvaradob-15616"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1174:is_spam", "suite": "spam", "family": "sms_phishing-3581acf70598e0b71c11", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 16 April, AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play. 8.Neighbour: We went 2 a marriage. Police arrested d murderer..who is it?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1174", "copies": 1, "upstream_ids": ["mendeley-1174"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4251:is_spam", "suite": "spam", "family": "sms_phishing-9f444c6826bc726e642e", "state": {"channel": "sms", "message": "I keep ten rs in my shelf to purchase a ball.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4251", "copies": 1, "upstream_ids": ["mendeley-4251"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2047:message_type", "suite": "spam", "family": "sms_phishing-4a0703262f126502e87f", "state": {"channel": "sms", "message": "Its ok chikku, and its my 1 of favourite song..:-)"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2047", "copies": 2, "upstream_ids": ["mendeley-2047", "uci-1775"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3911:is_spam", "suite": "spam", "family": "zefang_phishing_email-0c67b662b5971712c7b6", "state": {"channel": "email", "message": "Forwarded-by: Rob Windsor\nForwarded-by: \"David Dietz\" The latest proposal to drive the Taliban and Al Qaeda out of the\nMountains of Afghanistan is to send in the ASF (Alabama Special\nForces) Billy Bob, Bubba, Boo, Scooter, Cooter and Junior are being\nsent in with the following information about the Taliban:1. There is no limit.\n2. The season opened last weekend.\n3. They taste just like chicken.\n4. They hate beer, pickup trucks, country music, and Jesus.\n5. Some are queer.\n6. They don't like barbecue.And most importantly...7. They were responsible for Dale Earnhardt's death.We estimate it should be over in just about two days."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3911", "copies": 2, "upstream_ids": ["zefang-3911", "ealvaradob-3748"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2491:is_spam", "suite": "spam", "family": "sms_phishing-ca4dd1af95c9a8e60e3d", "state": {"channel": "sms", "message": "Captain vijaykanth is doing comedy in captain tv..he is drunken :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2491", "copies": 2, "upstream_ids": ["mendeley-2491", "uci-4639"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-655:message_type", "suite": "spam", "family": "sms_phishing-b096ad2db5a59a7361f3", "state": {"channel": "sms", "message": "Good FRIENDS CaRE for each Other.. CLoSE Friends UNDERSTaND each Other... and TRUE Friends STaY forever beyond words, beyond time. Gud ni8"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-655", "copies": 2, "upstream_ids": ["mendeley-655", "uci-4403"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00329-1:is_spam", "suite": "spam", "family": "generated_legit_email-71acd796577553acd219", "state": {"channel": "email", "message": "Gym Membership Renewal Reminder\n\nHi Christopher,\n\nPlease see the details below regarding your upcoming gym membership renewal at North Media Wellness Center. Your current term ends soon, and we want to ensure uninterrupted access to the facilities.\n\nMember Name: Christopher Lewis\nEmail: patrick.baker@web.de\nMembership ID: NMW-77392\nRenewal Date: 15 October 2022\nAmount Due: $65.00\nPlan Type: Premium Monthly\n\nTo complete your renewal, please log in to your member portal or visit the front desk with your employee badge by 14 October 2022. You can also renew online at https://www.northmedia.net/members/renew before the deadline to avoid a late processing fee.\n\nLet me know if you need help updating your payment method.\n\nBest regards,\nSusan Zhang\nFacilities Manager\nNorth Media Corporate Wellness"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00329-1", "copies": 1, "upstream_ids": ["gen-00329-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-1597:message_type", "suite": "spam", "family": "sms_phishing-9b2ed95d90d4455a79ba", "state": {"channel": "sms", "message": "Thank You meet you monday"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1597", "copies": 2, "upstream_ids": ["mendeley-1597", "uci-3546"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1410:is_spam", "suite": "spam", "family": "sms_phishing-5bbf989cfced7117e6c1", "state": {"channel": "sms", "message": "If i said anything wrong sorry de:-)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1410", "copies": 2, "upstream_ids": ["mendeley-1410", "uci-4413"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1773:is_spam", "suite": "spam", "family": "sms_phishing-d707d8a838692baa562f", "state": {"channel": "sms", "message": "Yun ah.now ü wkg where?btw if ü go nus sc. Ü wana specialise in wad?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1773", "copies": 2, "upstream_ids": ["mendeley-1773", "uci-1706"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81:is_spam", "suite": "spam", "family": "spamassassin-2bf50ce73d40ef98df77", "state": {"channel": "email", "message": "It's just too small HLOY\n\nA man endowed with a 7-8\" hammer is simply\n\nbetter equipped than a man with a 5-6\"hammer.\n\nWould you rather have\nmore than enough to get the job done or fall short. It's totally up\nto you. Our Methods are guaranteed to increase your size by 1-3\"\nCome in here and see how"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81", "copies": 5, "upstream_ids": ["20050311_spam_2/spam_2/00814.6b37fa2239e8c84e2237c4b156e16d81", "20050311_spam_2/spam_2/00825.decb3677f0081756d55f12d70b59a4e4", "20050311_spam_2/spam_2/00846.ed1959ae9b519ac6b944875b47497731", "20050311_spam_2/spam_2/00860.f1651a6a5f33bafe34e23afeacf85eb1", "20050311_spam_2/spam_2/01124.978b767962c91cc5c173cc4044b4658a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3169:is_spam", "suite": "spam", "family": "sms_phishing-06cf557b51885013ce43", "state": {"channel": "sms", "message": "Badrith is only for chennai:)i will surely pick for us:)no competition for him."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3169", "copies": 2, "upstream_ids": ["mendeley-3169", "uci-3281"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4957:message_type", "suite": "spam", "family": "sms_phishing-8479892da43d7268a46a", "state": {"channel": "sms", "message": "No worries, hope photo shoot went well. have a spiffing fun at workage."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4957", "copies": 2, "upstream_ids": ["mendeley-4957", "uci-3318"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3976:message_type", "suite": "spam", "family": "sms_phishing-4e250e43a2216597ee3b", "state": {"channel": "sms", "message": "You are being ripped off! Get your mobile content from www.clubmoby.com call 08717509990 poly/true/Pix/Ringtones/Games six downloads for only 3"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": [0.5, 0.0, 0.5], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3976", "copies": 3, "upstream_ids": ["mendeley-3976", "mendeley-4321", "uci-4643"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-10148:is_spam", "suite": "spam", "family": "zefang_phishing_email-83a3dfe6042c63adaefc", "state": {"channel": "email", "message": "new book : philosophy of language philosophy of language the big questions edited by andrea nye university of wisconsin , whitewater this wide-ranging anthology brings together a variety of readings in the philosophy of language from the ancient greeks to contemporary analytic , feminist , and multicultural perspectives . the emphasis is on issues that have a direct bearing on concerns about knowledge , reality , meaning , and understanding . a general introduction and introductions to each group of readings identify both the continuities and differences in the way \" big \" questions in philosophy of language have been addressed by philosophers of different historical periods , institutional affiliations , races , and genders . contents general introduction a : language : what is it ? : 1 . plato : language as the weaving together of forms 2 . jean - jacques rousseau : the origins of language 3 . john locke : on words 4 . gottlob frege : on the scientific justification of a conceptual notation 5 . john dewey : the existential matrix of inquiry 6 . ludwig wittgenstein : picturing reality 7 . emile benveniste : subjectivity in language 8 . jerry fodor : private language , public languages 9 . monique wittig : the mark of gender . b : meaning : how do words get their sense ? : 10 . gottlob frege : on sense and reference 11 . a . j . ayer : the principle of verifiability 12 . ludwig wittgenstein : meaning as use 13 . ruth millikan : biosemantics 14 . merrill and jaakko hintikka : how can language be sexist ? c : speaking : what is it to say something ? : 15 . h . p . grice : meaning 16 . john austin : performative utterances 17 . v . n . volosinov : verbal interaction 18 . arindam chakrabati : telling as letting know 19 . luce irigaray : he i sought but did not find . d : reference : what do we talk about ? : 20 . bertrand russell : on denoting 21 . saul kripke : naming and necessity ( excerpts ) 22 . michel foucault : the formation of objects 23 . judith butler : critically queer 24 . maria lugones : \" worlds \" and world travelling . e : truth : what is the relation between language and reality ? : 25 . alfred tarski : the semantic theory of truth 26 . donald davidson : the method of truth in metaphysics 27 . linda alcoff : truth as coherence 28 . david theo goldberg : truth through social praxis 29 . sandra harding : are truth claims dysfunctional ? f : other minds and foreign tongues : how is it possible to understand what someone else says ? : 30 . bronislaw malinowski : the translation of untranslatable words 31 . w . v . quine : indeterminacy of translation 32 . donald davidson : on the very idea of a conceptual scheme 33 . benjamin whorf : an american indian model of the universe 34 . gloria anzaldua : how to tame a wild tongue 35 . wole soyinka : language as boundary philosophy : the big questions series june 2020 6 x 9 336 pages 0-631 - 20602 - 7 paperback $ 29 . 95 0-631 - 20601 - 9 hardcover $ 62 . 95 available for review to order please call blackwell publishers at 1-800 - 216-2522 or visit our website at http : / / www . blackwellpub . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10148", "copies": 2, "upstream_ids": ["zefang-10148", "ealvaradob-9660"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5890:is_spam", "suite": "spam", "family": "sms_phishing-cc64af943d07b8293689", "state": {"channel": "sms", "message": "Double Mins & Double Txt & 1/2 price Call MobileUpd8 for the very latest offers. 08000839402 or call2optout/LF56"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5890", "copies": 1, "upstream_ids": ["mendeley-5890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00603-3:message_type", "suite": "spam", "family": "generated_legit_email-0874a38d976afc0dddc3", "state": {"channel": "email", "message": "Pioneer Health Monthly Newsletter - June 2013\n\nDear Timothy Walsh,\n\nThank you for subscribing to the Pioneer Health newsletter. We hope you are enjoying the start of summer! Below is your monthly health update and subscription information.\n\nSubscriber: Timothy Walsh\nAccount ID: PH-883021\nSubscription Type: Premium Wellness Digest\nBilling Cycle: Monthly\nNext Renewal Date: 2 July 2013\nAmount Charged: $12.00\n\nThis month, we are focusing on hydration and outdoor safety. Our featured article covers the best ways to stay cool during the summer heat without relying solely on sugary sports drinks. We also have a new guide on protecting your skin from UV damage, written by our resident dermatologist, Dr. Aris Thorne. Remember to reapply sunscreen every two hours if you are spending the day outside.\n\nAdditionally, we are excited to announce that our new patient wellness portal is officially live. You can now track your daily water intake, log your vitamin schedule, and message your care team directly through the dashboard. To explore all the new features and read this month's full articles, please visit: https://www.pioneerhealth.net/newsletter/june2013\n\nWe are also running a special promotion for our subscribers this month. Use code SUMMER08 at checkout to receive fifteen percent off all natural supplements in our online store. This offer expires on 30 June 2013, so be sure to stock up early.\n\nIf you ever wish to change your subscription preferences or unsubscribe entirely, you can manage your settings in your account dashboard. We value your privacy and will never share your personal information with third parties.\n\nWishing you a healthy and happy summer!\n\nWarm regards,\nKathleen Muller\nCommunity Outreach Director\nPioneer Health\nhttps://www.pioneerhealth.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00603-3", "copies": 1, "upstream_ids": ["gen-00603-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:nazario:none:phishing-2023#213:message_type", "suite": "spam", "family": "nazario-b54dd6b08af68d880c64", "state": {"channel": "email", "message": "Notice: Password Expiry Notice\n\nFrom Webmail willowlogistics.com Server Admin\n\nDear chloe\n\nTo help fight spam and improve security services, Please validate your mailbox\n\nValidate harbordesign.io\n\nSecure Mailbox\n\nThis email was sent to chloe to provide an update about your mailbox\n© 2012 | 7/26/2012 10:53:18 a.m."}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#213", "copies": 1, "upstream_ids": ["phishing-2023#213"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-4566:is_spam", "suite": "spam", "family": "zefang_phishing_email-ef2d24d6a66299ac9096", "state": {"channel": "email", "message": "start date : 1 / 23 / 02 ; hourahead hour : 15 ; start date : 1 / 23 / 02 ; hourahead hour : 15 ; no ancillary schedules awarded . no variances detected . log messages : parsing file - - > > o : \\ portland \\ westdesk \\ california scheduling \\ iso final schedules \\ 2002012315 . txt"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4566", "copies": 2, "upstream_ids": ["zefang-4566", "ealvaradob-4365"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1682:is_spam", "suite": "spam", "family": "sms_phishing-fb725ac16ace6b8c678e", "state": {"channel": "sms", "message": "We stopped to get ice cream and will go back after"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1682", "copies": 2, "upstream_ids": ["mendeley-1682", "uci-3739"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3484:message_type", "suite": "spam", "family": "sms_phishing-b3ba8096621960327809", "state": {"channel": "sms", "message": "Man this bus is so so so slow. I think you're gonna get there before me"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3484", "copies": 2, "upstream_ids": ["mendeley-3484", "uci-4495"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2539:is_spam", "suite": "spam", "family": "sms_phishing-6baf99e48dcf16b45b03", "state": {"channel": "sms", "message": "FREE>Ringtone! Reply REAL or POLY eg REAL1 1. PushButton 2. DontCha 3. BabyGoodbye 4. GoldDigger 5. WeBeBurnin 1st tone FREE and 6 more when u join for £3/wk"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2539", "copies": 2, "upstream_ids": ["mendeley-2539", "uci-3921"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00140-0:message_type", "suite": "spam", "family": "generated_unwanted_email-ce45ea8b60e4bdce738d", "state": {"channel": "email", "message": "Exclusive financial insights for Larry Muller\n\nDear Larry Muller,\n\nHarbor Consulting is pleased to share our latest quarterly wealth newsletter. We specialise in portfolio growth strategies and private banking advisory.\n\nOur early-bird consultation window closes on 31 May 2008. Please reply directly to this email to reserve your session with our senior advisors.\n\nWarm regards,\nCarolyn Hassan\nSenior Advisor, Harbor Consulting"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00140-0", "copies": 1, "upstream_ids": ["genu-00140-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited marketing newsletter from a shop the reader never signed up to", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-477:message_type", "suite": "spam", "family": "sms_phishing-ff943112f6446be60ec9", "state": {"channel": "sms", "message": "LookAtMe!: Thanks for your purchase of a video clip from LookAtMe!, you've been charged 35p. Think you can do better? Why not send a video in a MMSto 32323."}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-477", "copies": 2, "upstream_ids": ["mendeley-477", "uci-3132"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4489:message_type", "suite": "spam", "family": "sms_phishing-d0ffc9ab2683e90ba09f", "state": {"channel": "sms", "message": "I dont understand your message."}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4489", "copies": 2, "upstream_ids": ["mendeley-4489", "uci-2874"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-919:is_spam", "suite": "spam", "family": "sms_phishing-257bbaf0164d334d1f31", "state": {"channel": "sms", "message": "I.ll hand her my phone to chat wit u"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-919", "copies": 2, "upstream_ids": ["mendeley-919", "uci-5300"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-11947:is_spam", "suite": "spam", "family": "zefang_phishing_email-a095e6c73c7cc35f6fc2", "state": {"channel": "email", "message": "pre _ ejaculat ; ors : please come here for info on a recent medical miracle . not interested tap here"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11947", "copies": 2, "upstream_ids": ["zefang-11947", "ealvaradob-11331"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17238:is_spam", "suite": "spam", "family": "zefang_phishing_email-f1e723152972715f30fd", "state": {"channel": "email", "message": "Ned Jackson Lovely wrote:\nneed to\nthe tree lovers lawsuits that prevented removal of trees by any means as at\nleast partially responsible. The lawsuits prevented the Forest Service\nfrom implementation of any project that involved tree removal. These\nprojects were designed to address the tree and undergrowth density that\nresulted from previous management policies.In other words, Forest Service projects to clean up the mess resulting from\ntheir previous forest management practices are prevented by tree lover\nlawsuits. > 40 years after the end of the containment policy there is still a lot of\nstuff\nstuff. The result: Fires like Cerro Grande. > I can't imagine that artificially planted replacement\nare mostly planted on private land as commercial tree farms.Peace,Marty Halvorson"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17238", "copies": 2, "upstream_ids": ["zefang-17238", "ealvaradob-16237"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb:is_spam", "suite": "spam", "family": "spamassassin-ee80339c087d97b78258", "state": {"channel": "email", "message": "Martin Adamson wrote:\n\nYou're probably right, mine's just the same but with the cream added to the\neggs. I guess I should try it without. Actually looking on the internet for a\nrecipe I found this one from possibly one of the scariest people I've ever seen,\nand he's a US Congressman:\n<http://www.virtualcities.com/ons/me/gov/megvjb1.htm>\n\nThat's one of the worst non-smiles ever.\n\nStew\nps. Apologies if any of the list's Maine residents voted for this man, you won't\ndo it again once you've seen this pic.\n\n--\nStewart Smith\nScottish Microelectronics Centre, University of Edinburgh.\nhttp://www.ee.ed.ac.uk/~sxs/\n\n4 DVDs Free +s&p Join Now\nhttp://us.click.yahoo.com/pt6YBB/NXiEAA/mG3HAA/7gSolB/TM"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2484:message_type", "suite": "spam", "family": "sms_phishing-31390e4f0a5a59c9cc7a", "state": {"channel": "sms", "message": "Just now saw your message.it k da:)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2484", "copies": 2, "upstream_ids": ["mendeley-2484", "uci-4426"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2915:is_spam", "suite": "spam", "family": "sms_phishing-c04db586ca83488df172", "state": {"channel": "sms", "message": "PRIVATE! Your 2013 Account Statement for 078"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2915", "copies": 2, "upstream_ids": ["mendeley-2915", "uci-5122"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00256-3:message_type", "suite": "spam", "family": "generated_unwanted_email-09d03cc15b8a8096d158", "state": {"channel": "email", "message": "Action Required: Customs Clearance Fee for Delivery\n\nDear Anna Nguyen,\n\nWe are contacting you regarding a parcel currently held at our regional distribution centre. The package could not be delivered because a small customs clearance fee remains outstanding. As Atlas Works Financial Services manages the escrow payments for this shipment, we have been instructed to collect the balance before the courier can proceed.\n\nPlease review the shipment details below and settle the fee so that your delivery can be released without further delay.\n\nRecipient: Anna Nguyen\nEmail on file: dennis.cooper@aol.com\nTracking Reference: AW-992810447\nCarrier: Atlas Express Logistics\nDelivery Attempt Date: 25 April 2008\nStatus: Held at Customs\nAmount Due: GBP 3.45\nPayment Deadline: 30 April 2008\n\nTo ensure the security of your transaction, the payment must be made using a valid debit or credit card through our encrypted portal. Once the fee has been successfully processed, your parcel will be dispatched for redelivery within twenty-four hours.\n\nPlease click the link below to verify your card details and pay the outstanding amount:\nhttps://www.atlasworks-secure.co.uk/payments/customs-clearance/AW-992810447\n\nIf the fee is not settled by the deadline shown above, the parcel will be returned to the sender, and additional return shipping charges may apply. We understand that unexpected fees can be inconvenient, but this charge is required by international shipping regulations for goods entering the country.\n\nShould you experience any difficulty completing the payment online, please reply to this message and our support team will be happy to assist you.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nGary James\nPayment Resolution Team\nAtlas Works Financial Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00256-3", "copies": 1, "upstream_ids": ["genu-00256-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00026-1:message_type", "suite": "spam", "family": "generated_unwanted_email-5675c02feb304c0e0790", "state": {"channel": "email", "message": "Unusual login attempt blocked\n\nDiego,\n\nYour account was locked after a sign-in from Lagos. Confirm your details now to unlock it: https://www.meridianmedia-secure.io/verify-diego\n\nDaniel Hassan\nIT Security, Meridian Media"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00026-1", "copies": 1, "upstream_ids": ["genu-00026-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00259-3:is_spam", "suite": "spam", "family": "generated_legit_email-2298a6906883d025bb04", "state": {"channel": "email", "message": "Your ticket confirmation\n\nHi Heather,\n\nJust confirming that your ticket for the Blue Labs Annual Developer Summit on August 14, 2023, has been booked. Your registration number is BL-9932. Please ensure you bring a valid photo ID to the registration desk before 8:30 AM on the day of the event.\n\nBest,\nWilliam Patel\nBlue Labs Registration Team"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00259-3", "copies": 1, "upstream_ids": ["gen-00259-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:nazario:none:phishing-2023#188:message_type", "suite": "spam", "family": "nazario-1f52725a3938f1cbd3e4", "state": {"channel": "email", "message": "Please Reset your information #300929\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 42789220This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#188", "copies": 1, "upstream_ids": ["phishing-2023#188"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d:message_type", "suite": "spam", "family": "spamassassin-3638619e0c52ccda8f04", "state": {"channel": "email", "message": "spamd can't find...\n\nI thought I'd installed razor correctly, but I am seeing the following in my\nlogs. Can anyone give me a hitn?\n\nSep 8 06:46:45 omega spamd[6514]: razor2 check skipped: No such file or\ndirectory Can't locate object method \"new\" via package \"Net::DNS::Resolver\"\n(perhaps you forgot to load \"Net::DNS::Resolver\"?) at (eval 31) line 1,\n<STDIN> line 114. ^I...propagated at\n/usr/lib/perl5/site_perl/5.6.1/Mail/SpamAssassin/Dns.pm line 392, <STDIN>\nline 114.\n\n--\nRobin Lynn Frank\nParadigm-Omega, LLC\nThe only certainty about documentation is that\nwhoever wrote it \"might\" have understood it.\nThe rest of us may not be so lucky.\n\ncell phone? Get a new here for FREE!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00709-1:message_type", "suite": "spam", "family": "generated_legit_email-d43adccb84dbe27e0a51", "state": {"channel": "email", "message": "Spring Training Schedule and Membership Fees\n\nHi Kenji,\n\nHope you are doing well! I am reaching out with a quick update regarding the Westfield Athletics Club spring training schedule. As the weather improves, we are moving our sessions back outdoors starting Monday 24 April 2011.\n\nTraining will now take place every Monday and Wednesday evening from 6:30 PM to 8:00 PM at Memorial Park, pitch 3. Saturday morning fixtures remain unchanged for now, but I will send another note once the May league matches are confirmed.\n\nPlease remember that the annual membership renewal fee of £45.00 is due by Friday 28 April 2011. You can pay this by bank transfer or drop cash off with me at Wednesday's session. We cannot register athletes for the upcoming county meet without cleared dues.\n\nLet me know if you have any trouble making the new times work with your schedule.\n\nBest wishes,\nDebra Young\nClub Secretary\nDeltalabs Community Sports Association"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00709-1", "copies": 1, "upstream_ids": ["gen-00709-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "sports or hobby club message (training times, fees, fixtures)"}}
{"id": "spam:zefang_phishing_email:none:zefang-6849:is_spam", "suite": "spam", "family": "zefang_phishing_email-f9cd159896c3e3452c78", "state": {"channel": "email", "message": "psycholinguistics jean aitchison , the articulate mammal an introduction to psycholinguistics , 4th edition this highly successful text has been substantially revised to take account of the considerable changes in chomsky 's recent ideas . the chapters on grammatical innateness , child language acquisition , and speech comprehension have been largely rewritten and new material and references throughout the book make it completely up-to - date with key developments in the field . table of contents introduction 1 . the great automatic grammatizor : need anything be innate ? 2 . animals that try to talk : is language restricted to humans ? 3 . grandmama 's teeth : is there biological evidence for innate language capacity ? 4 . predestinate grooves : is there a pre-ordained language ' programme ' ? 5 . the blueprint in the brain : what grammatical information might conceivably be innate ? 6 . chattering children : are chattering children following ` rules ' when they learn to speak ? 7 . puzzling it out : exactly how do children learn language ? 8 . celestial unintelligibility : why propose a transformational grammar ? 9 . the white elephant problem : do we need a transformational grammar in order to speak ? 10 . the case of the missing fingerprint : how do we understand speech ? 11 . the cheshire cat 's grin : how do we plan and produce speech ? 12 . banker 's clerk or hippopotamus ? : the future of psycholinguistics notes and suggestions for further reading references routledge : 2023 : 320 pp cl : 0 415 16866 x : # d5684 : $ 75 . 00 pb : 0 415 16791 4 : # d5688 : $ 24 . 99 for more information on these and other titles from : routledge london * new york in north america : www . routledge-ny . com elsewhere : www . routledge . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6849", "copies": 2, "upstream_ids": ["zefang-6849", "ealvaradob-6563"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1429:message_type", "suite": "spam", "family": "nazario-b1350118fa1d6a9ba34a", "state": {"channel": "email", "message": "Regions Bank Account Maintenance\n\nDear ohoward@icloud.com,\nREGIONS BANK CUSTOMER\n\nIt has come to our attention that your Regions Bank Account Maintenance tax of USD 0.95\nare now overdue. These fees result from using related services:\nAccount Alerts, Online Banking, Online Statements, Reorder Checks, Stop Payment.\n\nTo prevent any disruption in your ability to use your account, please pay your Regions Bank fees by following these steps:\n\n1. Log in immediately using the link http://0x45.0x02.0x43.0xf3/regionsnet/EB/logon/\n2. Complete the mandatory fields.\n3. Done.\n\nPlease update your payment information. We apologize for the inconvenience\nAt Regions, we are committed to safeguarding your personal information, whether it is in a branch or\non one of our Web sites.\n\nSincerely,\nCarter Franke\nChief Marketing Officer\nRegions Bank Customer Care - a CUNA.org\n\nPROTECT YOUR PASSWORD\n\nNEVER give your password to anyone\nProtect yourself against fraudulent websites by\nopening a new web browser (e.g. Internet Explorer or Netscape) and typing\nin the URL every time you log in to your account.\n\nPlease do not reply to this e-mail. Mail sent to this address cannot be\nanswered. For assistance, log in to your Regions-Net account and choose the\n\"Help\" link in the header of any page.\n\nEmail ID REGNET321"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1429", "copies": 1, "upstream_ids": ["phishing3.mbox#1429"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3580:message_type", "suite": "spam", "family": "sms_phishing-d16e554a7cfa8672dcf1", "state": {"channel": "sms", "message": "No shit, but I wasn't that surprised, so I went and spent the evening with that french guy I met in town here and we fooled around a bit"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3580", "copies": 1, "upstream_ids": ["mendeley-3580"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4473:is_spam", "suite": "spam", "family": "sms_phishing-9f33f8dd3f74396fbdc5", "state": {"channel": "sms", "message": "No dear. . Tanu going to talk in surya tv"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4473", "copies": 1, "upstream_ids": ["mendeley-4473"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1820:message_type", "suite": "spam", "family": "sms_phishing-110c01e60431281869ad", "state": {"channel": "sms", "message": "You have been selected to stay in 1 of 250 top British hotels - FOR NOTHING! Holiday Worth £350! To Claim, Call London 02072069400. Bx 526, SW73SS"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1820", "copies": 2, "upstream_ids": ["mendeley-1820", "uci-1628"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#1018:message_type", "suite": "spam", "family": "nazario-62962ddfc18e075562eb", "state": {"channel": "email", "message": "Your account access has been limited\n\nHSBC Bank - Limited Account Access Details\n\nÂ\n\nHSBC Bank is constantly working to ensure security by regularly screening the accounts in our system. We recently reviewed your account, and we need more information to help us provide you with secure service. Until we can collect this information, your access to sensitive account features will be limited. We would like to restore your access as soon as possible, and we apologize for the inconvenience.\n\nWhy is my account access limited?\n\nYour account access has been limited for the following reason(s):\n\nJun. 2, 2012: We have detected suspicious activity regarding the receipt or withdrawal of funds.\n\n(Your case ID for this reason is PP-171-147-406.)\n\nHow can I restore my account access?\n\nPlease visit the Resolution Center and complete the \"Steps to Remove Limitations.\"\n\nOnce you complete all of the checklist items, your case will be reviewed by one of our Account Specialists. We will send you an email with the outcome of the review."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1018", "copies": 1, "upstream_ids": ["phishing2.mbox#1018"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3036:is_spam", "suite": "spam", "family": "sms_phishing-892278952419658b3582", "state": {"channel": "sms", "message": "Hey I am really * want to chat or see me text to 08718738034"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3036", "copies": 1, "upstream_ids": ["mendeley-3036"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-11449:is_spam", "suite": "spam", "family": "zefang_phishing_email-69700ada6a109ceb8b5b", "state": {"channel": "email", "message": "black marlin have you had a chance to look at this yet ? - - - - - original message - - - - - from : schumack , sherlyn sent : friday , september 07 , 2020 4 : 41 pm to : olsen , michael cc : farmer , daren j . subject : black marlin mike , i show you pathed deal 951474 on 8 / 23 / 01 for production months january and february 2020 . this deal expired 12 / 31 / 00 , and the late pathing has caused a bridgeback error for january and february 2020 ."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11449", "copies": 2, "upstream_ids": ["zefang-11449", "ealvaradob-10870"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2:is_spam", "suite": "spam", "family": "spamassassin-d223622360a78d9332ea", "state": {"channel": "email", "message": "What you need for iPaq wireless connectivity (Tech Update)\n\nTech Update Today\n\nVITAL SIGNS FOR JULY 12, 2014\n\nDavid Berlind\n\nWhat you need for iPaq wireless connectivity\n\nA public health organization has several iPaqs on order, but\nnow they want to add WAN connectivity to those PDAs. What\nexactly do they need to do? I'm so glad they asked.\n\nHere's everything you need to know\n\nLatest from ZDNet News\n\nSecurity holds back Web services\n\nChina: Bursting with brainpower\n\nMicron cranks handle on next-gen memory\n\nJapan chipmakers team on 'SoCs' chip\n\nHP tech exec jumps ship to EMC\n\nHoles put Linux handheld at risk\n\nMicrosoft bids to be creme de la CRM\n\nMore Enterprise News\n\nFarber's Picks\n\nDan\nFarber\n\nDesperately seeking chief ethics officer\n\nGreedy CEOs are making a bad name for the profession--and corporate America. The title chief executive officer now\nconjures up an image of someone with questionable ethics, a\ngreedy disposition, and a disregard for shareholders. If\ngovernment-proposed remedies can\"t adequately address this\nethics crisis, disgruntled shareholders surely will.\n\nRead my full commentary\n\nOracle challenges Microsoft anew\n\nOracle is taking on Microsoft's Exchange with a Collaboration Suite that includes e-mail, calendar and messaging software. Oracle has pitched its products as Microsoft replacements before -- with spectacularly little success. (At Comdex last November, the database giant offered its 9i and app-server software as a more secure alternative to Exchange.) What's different this time? Oracle is selling a separate collaboration product that, says CEO Larry Ellison, is cheaper than rival offerings. Plus: Companies facing a deadline to join Redmond's new licensing program may be more open to non-Micrososft alternatives.\n\nRead the full story\n\nAlso: Ellison shrugs a No. 2 exec\n\nStandards stalled over royalty disputes\n\nThe World Wide Web Consortium is bracing for a vote next week that could decide once and for all how the standards body will handle patented technology that comes with royalties attached. If patented technology is rejected, the vote could mark a major shift in the Internet standards battle. Major intellectual property holders such as Microsoft and IBM have a lot at stake.\n\nRead the full story\n\nMicrosoft's OpenGL claim shakes graphics world\n\nSurvey: Execs rethinking IT budgets\n\nGrowing numbers of Fortune 1000 CIOs are re-evaluating their IT budgets and spending plans for coming months, according to a recent survey. Morgan Stanley finds that CIOs' top priorities include security software, application integration and Windows 2012/XP upgrades for the desktop. Integration challenges were slowing the rate of application adoption for 38 percent of those surveyed.\n\nRead more survey results\n\nOh look, a new bug in Outlook and IE\n\nA Danish security expert has warned of a security flaw in Microsoft's ActiveX scripting component that allows network attackers to peek at cookies, read files and even run programs on a victim's PC. The expert recommended that users disable ActiveX in IE and Outlook, at least until Microsoft releases a patch.\n\nRead the full story\n\nAlso: Outlook plug-in opens PCs to hackers\n\nWrite me at sofia_brown@orange.fr\n\nBack to top\n\nAlso on Tech Update Today\n\nCOMMENTARY\n\nMixing Unix and Windows\n\nMost enterprises have a mix of Windows and Unix systems. Larry Seltzer explores the options for interoperating those systems.\n\nQ&A\n\nMarcelo Tosatti: The future is Linux\n\nThe maintainer of the Linux kernel talks about new opportunities, defends UnitedLinux and tells why copyright laws keep him from entering the U.S.\n\nPREVIOUSLY ON TECH UPDATE TODAY\n\nROI calculators don't tell the whole TCO story\n\nIn response to David Berlind's \"Dell's formula for ROI,\" TCO\nconsultant Tom Rankin warns that \"99 percent of TCO tools do\nnot look at the 'total' but may look at whatever parameters\nsuit the vendor to sell their products/services.\"\n\nCrucial Clicks--products worth looking at\n\nNetworking\n\nFind the right home network\nChoosing the right technology to match your networking needs is vital. Here's help in making the choice.\n\nRead reviews\n\nMost Popular Products\n\nNetworking\n\n1. Linksys EtherFast wireless AP\n\n2. Linksys EtherFast router\n\n3. Siemens SpeedStream router\n\n4. Wireless 802.11B Router\n\n5. NetGear HE102 802.11a wireless AP\n\nMore popular networking products\n\nElsewhere on ZDNet\n\nVeritas CEO Gary Bloom speaks out at Tech Update.\nCheck out these personal laser printer picks at ZDNet Reviews.\nSee how WAP works and how you should use it at Builder.com.\nNeed a new job? Find one today in ZDNet's Career Center.\nCIOs talk out about the future of IT at CNET News.com.\n\nSign up for more free newsletters from ZDNet\n\nThe e-mail address for your subscription is tbrown@orange.fr\n\nManage\nMy Subscriptions | FAQ |\nAdvertise\n\nHome |eBusiness | Security | Networking | Applications | Platforms | Hardware | Contact us\n\nCopyright 2014 CNET Networks, Inc. All rights reserved. ZDNet is a registered service mark of CNET Networks, Inc."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-416:is_spam", "suite": "spam", "family": "sms_phishing-900c43c65f076cf807d8", "state": {"channel": "sms", "message": "Wen did you get so spiritual and deep. That's great"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-416", "copies": 2, "upstream_ids": ["mendeley-416", "uci-5553"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00763-1:is_spam", "suite": "spam", "family": "generated_legit_email-0ee0dc0d72a3bd6d8340", "state": {"channel": "email", "message": "Membership Renewal Reminder - Westside Community Garden\n\nDear Hans,\n\nThis is a friendly reminder that your annual membership with the Westside Community Garden is set to expire on March 31, 2014. To keep your assigned plot active and maintain access to the tool shed, please renew your membership before the deadline.\n\nThe renewal fee remains $45.00 for the year. You can pay by mailing a check to our office or by dropping off cash during our weekend volunteer hours on Saturday mornings.\n\nIf you choose not to renew, your plot will be reassigned to someone on our waiting list starting April 1. We would hate to lose you after such a great season last year!\n\nIf you have any questions or need assistance with the renewal process, please give me a call at (504) 555-0172.\n\nWarmly,\nKenneth Moreau\nMembership Coordinator\nWestside Community Garden"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00763-1", "copies": 1, "upstream_ids": ["gen-00763-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-5173:message_type", "suite": "spam", "family": "sms_phishing-135fdbb45a0014dc442d", "state": {"channel": "sms", "message": "JADE ITS PAUL. Y DIDN’T U TXT ME? DO U REMEMBER ME FROM BARMED? I WANT 2 TALK 2 U! TXT ME"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5173", "copies": 2, "upstream_ids": ["mendeley-5173", "uci-4979"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3992:is_spam", "suite": "spam", "family": "sms_phishing-8bdb60f7e06087b71bd4", "state": {"channel": "sms", "message": "Do you know when dad will be back?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3992", "copies": 2, "upstream_ids": ["mendeley-3992", "uci-2662"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5916:message_type", "suite": "spam", "family": "sms_phishing-351b6aa3a9eb0386324b", "state": {"channel": "sms", "message": "Had your mobile 10 mths? Update to the latest Camera/Video phones for FREE. KEEP UR SAME NUMBER, Get extra free mins/texts. Text YES for a call"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5916", "copies": 2, "upstream_ids": ["mendeley-5916", "uci-4728"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1702:is_spam", "suite": "spam", "family": "sms_phishing-0f06478b84026de5633a", "state": {"channel": "sms", "message": "todays vodafone numbers ending with 0089(my last four digits) are selected to received a £350 award. If your number matches please call 09063442151 to claim your £350 award"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1702", "copies": 2, "upstream_ids": ["mendeley-1702", "uci-1687"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2350:is_spam", "suite": "spam", "family": "sms_phishing-e55ba989d091a1d148eb", "state": {"channel": "sms", "message": "IMPORTANT INFORMATION 4 ORANGE USER 0796XXXXXX. TODAY IS UR LUCKY DAY!2 FIND OUT WHY LOG ONTO http://www.urawinner.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2350", "copies": 1, "upstream_ids": ["mendeley-2350"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-577:is_spam", "suite": "spam", "family": "zefang_phishing_email-aab708c168d1145e452c", "state": {"channel": "email", "message": "journal of cognitive systems research call for papers : new electronic journal of cognitive systems research call for papers journal of cognitive systems research editors - in - chief ron sun e - mail : rsun @ cs . ua . edu department of computer science and department of psychology university of alabama tuscaloosa al , usa vasant honavar e - mail : honavar @ cs . iastate . edu department of computer science iowa state university usa gregg oden e - mail : gregg-oden @ uiowa . edu department of psychology university of iowa usa the journal of cognitive systems research covers all topics in the study of cognitive processes , in both natural and artificial systems : knowledge representation and reasoning learning perception action memory problem - solving and cognitive skills language and communication agents integrative studies of cognitive systems the journal emphasizes the integration / synthesis of ideas , concepts , constructs , theories , and techniques from multiple paradigms , perspectives , and disciplines , in the analysis , understanding and design of cognitive and intelligent systems . contributions describing results obtained within the traditional disciplines are also sought if such work has broader implications and relevance . the journal seeks to foster and promote the discussion of novel approaches in studying cognitive and intelligent systems . it also encourages cross-fertilization of disciplines , by publishing high-quality contributions in all of the areas of study , including artificial intelligence , linguistics , psychology , psychiatry , philosophy , system and control theory , anthropology , sociology , biological sciences , and neuroscience . the scope of the journal includes the study of a variety of different cognitive systems , at different levels , ranging from social / cultural cognition , to individual cognitive agents , to components of such systems . of particular interest are theoretical , experimental , and integrative studies and computational modeling of cognitive systems , at different levels of detail , and from different perspectives . please send submissions in postscript format by electronic mail to one of the three co - editors - in - chief . note the journal transends traditional disciplinary boundaries , and considers contributions from all relevant disciplines and approaches . the key is the quality of the work and the accessibility and relevance to readers in different disciplines . the first issue of this new on-line journal , published by elsevierscience , will appear in early 2023 . in addition to this electronic journal , the issues will also be printed and bound as archival volume . published papers will be considered automatically for inclusion in specially edited books on cognitive systems research . for further information , see : http : / / cs . ua . edu / ~ rsun / journal . html - - - - - - - - - - - - - - - - - - - action editors : john barnden , school of computer science , university of birmingham , u . k . \\ \\ william bechtel , department of philosophy , washington university , st . louis , usa . \\ \\ rik belew , computer science and engineering department , university of california , san diego , usa . \\ \\ mark h . bickhard , department of psychology , lehigh university , usa . \\ \\ deric bownds , dept . of zoology , university of wisconsin , madison , usa . \\ \\ david chalmers , department of philosophy , university of california , santa cruz , usa . \\ \\ b . chandrasekaran , department of computer and information science , ohio state university , usa . \\ \\ marco dorigo , university of brussels , brussels , belgium \\ \\ michael dyer , computer science department , university of california , los angeles , usa . \\ \\ lee giles , nec research institute , princeton , new jersey , usa . \\ \\ george graham , philosophy department , university of alabama at birmingham , birmingham , al , usa . \\ \\ stephen j . hanson , psychology dept . , rutgers university , newark , new jersey , usa . \\ \\ valerie gray hardcastle , dept . of philosophy , virginia polytechnic and state university , blacksburg , virginia , usa . \\ \\ james hendler , department of computer science , university of maryland , college park , usa . \\ \\ stephen m . kosslyn , department of psychology , harvard university , usa . \\ \\ george lakoff , dept . of linguistics , university of california , berkeley , usa . \\ \\ joseph ledoux , center for neuroscience , new york university , new york , usa . \\ \\ daniel levine , department of psychology , university of texas at arlington , usa . \\ \\ vladimir j . lumelsky , robotics laboratory , department of mechanical engineering , university of wisconsin , madison , usa . \\ \\ james pustejovsky , brandeis university , massachusetts , usa . \\ \\ lynne m . reder , department of psychology , carnegie mellon university , pittsburgh , pa 15213 , usa . \\ \\ jude shavlik , computer sciences department , university of wisconsin , madison , usa . \\ \\ tim shallice , department of psychology , university college , london , uk \\ \\ aaron sloman , school of computer science , the university of birmingham , uk . \\ \\ paul thagard , philosophy department , university of waterloo , canada . \\ \\ leonard uhr , computer sciences department , university of wisconsin , madison , usa . \\ \\ david waltz , nec research institute , princeton , nj , usa . \\ \\ xin yao , dept . of computer science , australian defense force academy , canberra , australia . \\ \\"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-577", "copies": 2, "upstream_ids": ["zefang-577", "ealvaradob-558"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00230-4:message_type", "suite": "spam", "family": "generated_legit_email-3b39ec6e2a8d31e8513a", "state": {"channel": "email", "message": "Reminder: Utility bill due next week\n\nHi Timothy,\n\nJust a quick reminder that your February utility bill of $142.50 is due on 15 February 2022. Please arrange payment before the deadline to avoid a late fee.\n\nIf you have any questions, give us a call at (555) 812-3304.\n\nThanks,\nFatima White\nOak Partners"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00230-4", "copies": 1, "upstream_ids": ["gen-00230-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "utility or phone bill reminder with a real due date"}}
{"id": "spam:sms_phishing:none:mendeley-2331:message_type", "suite": "spam", "family": "sms_phishing-8d9e7f4ecb8bb3eb370e", "state": {"channel": "sms", "message": "Doing nothing, then u not having dinner w us?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2331", "copies": 2, "upstream_ids": ["mendeley-2331", "uci-4379"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1721:is_spam", "suite": "spam", "family": "sms_phishing-5ea684ab791c6b45f0b0", "state": {"channel": "sms", "message": "Dont put your phone on silent mode ok"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1721", "copies": 2, "upstream_ids": ["mendeley-1721", "uci-1278"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2571:is_spam", "suite": "spam", "family": "sms_phishing-42c6a8959dfb93ba82b9", "state": {"channel": "sms", "message": "Come to me right now, Ahmad"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2571", "copies": 2, "upstream_ids": ["mendeley-2571", "uci-2412"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2646:message_type", "suite": "spam", "family": "sms_phishing-82a7ffb65c1aeca5cf9b", "state": {"channel": "sms", "message": "Nah it's straight, if you can just bring bud or drinks or something that's actually a little more useful than straight cash"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2646", "copies": 2, "upstream_ids": ["mendeley-2646", "uci-625"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00646-0:message_type", "suite": "spam", "family": "generated_legit_email-dcac9b4bdff92927d567", "state": {"channel": "email", "message": "Interview Invitation – Content Strategist Position\n\nDear Pamela Nelson,\n\nThank you for your recent application to North Media Group for the Content Strategist position. We have reviewed your materials and were impressed by your background in digital publishing. We would like to invite you to an interview to discuss the role further.\n\nPlease review the following details regarding your upcoming interview:\n\n1. Interview Date: Wednesday, 12 June 2008\n2. Interview Time: 10:30 AM (BST)\n3. Location: North Media Group Headquarters, 45 Fleet Street, London EC4Y 1DP\n4. Interviewer: Noah Byrne, Head of Editorial Operations\n\nTo ensure the process runs smoothly, please complete the numbered steps below before your arrival:\n\n1. Confirm your attendance by replying directly to this email no later than Friday, 7 June 2008.\n2. Prepare a brief portfolio presentation (maximum ten minutes) highlighting three campaigns you have managed.\n3. Bring two forms of photographic identification for our building security check-in.\n4. Arrive fifteen minutes early so we can offer you tea or coffee before the meeting begins.\n\nDuring the interview, we will cover the following topics:\n- Your experience with SEO-driven content planning\n- Team leadership and cross-department collaboration\n- Salary expectations and your earliest available start date\n\nWe are very much looking forward to meeting you in person and learning more about how your skills align with our current projects at North Media Group. If you require any reasonable adjustments for the interview or need directions to our office, please do not hesitate to reach out.\n\nYou can contact me directly by phone at +44 20 7946 0123 during standard business hours (Monday to Friday, 9:00 AM to 5:00 PM).\n\nWarm regards,\nNoah Byrne\nHead of Editorial Operations\nNorth Media Group\n+44 20 7946 0123"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00646-0", "copies": 1, "upstream_ids": ["gen-00646-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-1315:message_type", "suite": "spam", "family": "sms_phishing-8fd164d5fe8076775140", "state": {"channel": "sms", "message": "Will you come online today night"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1315", "copies": 2, "upstream_ids": ["mendeley-1315", "uci-4907"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6496:is_spam", "suite": "spam", "family": "zefang_phishing_email-b38effc7aa930b5a0e37", "state": {"channel": "email", "message": "accomplishments attached is a summary of accomplishments for first half as requested ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6496", "copies": 2, "upstream_ids": ["zefang-6496", "ealvaradob-6227"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-14010:is_spam", "suite": "spam", "family": "zefang_phishing_email-57fc768e92aa1f9933ac", "state": {"channel": "email", "message": "cialis ( super viagra ) at $ 2 . 00 per dose ready to boost your sex life ? positive ? time to do it right now . order cialis at incredibly low prices $ 2 . 00 per dose . unbelivable no more"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14010", "copies": 2, "upstream_ids": ["zefang-14010", "ealvaradob-13251"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2664:is_spam", "suite": "spam", "family": "sms_phishing-6c4f0a6b26a8a43f8749", "state": {"channel": "sms", "message": "Hi frnd, which is best way to avoid missunderstding wit our beloved one's?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2664", "copies": 2, "upstream_ids": ["mendeley-2664", "uci-168"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00068-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-2b4376ca8a8324b631e7", "state": {"channel": "email", "message": "Mailbox storage limit reached\n\nSamantha,\n\nYour mailbox has exceeded the 2 GB storage quota as of this morning. Incoming messages will be rejected until you verify your account credentials to authorize an automatic archive expansion.\n\nPlease reply with your current password so IT can process the upgrade before end of day.\n\nRegards,\nNancy Bennett\nIT Support Desk\nCedar Design Ltd"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00068-4", "copies": 1, "upstream_ids": ["genu-00068-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-11840:is_spam", "suite": "spam", "family": "zefang_phishing_email-a6320ab2b8a5cfc64981", "state": {"channel": "email", "message": "correction to call for abstracts the call for abstracts posted earlier gave the deadline for submissions as march , 2020 . please send them this year . the corrected call is as follows : call for abstracts american dialect society , midwest regional meeting submission deadline : march 27 , 2021 , for american dialect society session midwest modern language association november 2 - 4 , 2021 st . louis , mo marriott pavilion hotel abstracts and proposals on any topic welcome . if there is sufficient interest , we will have a session devoted to papers , presentations and discussion on topics related to midland dialect . send to beth lee simon , simon @ cvax . ipfw . indiana . edu cm 109 department of english and linguistics indiana university - purdue university at fort wayne fort wayne , in 46805 * * * * * * * * * * * * * * * * * * * * *"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11840", "copies": 2, "upstream_ids": ["zefang-11840", "ealvaradob-11230"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3:message_type", "suite": "spam", "family": "spamassassin-ca1929840c51235da871", "state": {"channel": "email", "message": "The Curse of India's Socialism\n\nThe Philippines suffers from a Spanish-imposed aristocratic landowner\nclass, the same as in most of Latin America, with the serf-like\nrelationships that go with it. You have multiple generations of\npeasants/squatters that cultivate and live on the lands almost as a\nhuman parts of the property package.\n\nThe problem in the Philippines is that the landowner class is fading,\nbut the peasants are still operating under the old economic assumption\neven though it isn't being imposed on them in most cases. In a number\nof areas, you have landowners that don't use their lands during bad\nyears (i.e. when sugar prices fall too low or some other condition that\nmakes planting unprofitable in the global economy) and in some cases\nhave all but abandoned them, but when they come back to plant crops\nseveral years later they find the same peasants living on the fallow\nland as scratch farmers to survive and waiting for the landowner to come\nback and give them something to do.\n\nThe government's solution to this was to \"redistribute\" fallow land to\nthe peasants that lived on them, and created a number of problems in the\nprocess. First of all, it meant that all the agricultural producers had\nto plant crops all the time (profitable or not) or the government would\nseize the land and give it to someone else, which is not economically\noptimal by any means. Second when they give it to the peasants, they\nuse it to become scratch farmers, which is not a particularly productive\neconomic activity, and the landowners generally let them do that for\nfree when the land is fallow anyway. Third, the peasants generally\n*prefer* to work for the landowner rather than own the land themselves;\nthey make more money, have more opportunities to become middle-class,\nand enjoy a better lifestyle under that arrangement. So the result of\nall this is that the landowners either run an economically sub-optimal\nagriculture business to protect their lands, or they risk having their\nland seized without compensation and given to the peasants/squatters who\nwill gladly sell it back to the plantation owner rather than becoming\nscratch farmers, another economically sub-optimal solution. The only\nreal role of the government in this arrangement is to be a drag on the\neconomy.\n\nAnd I don't know about 20 years for land titles in the Philippines. It\ntakes a little longer than in the US (though not much), and you may have\nto bribe someone to make it work well in an expedited manner, but it\nisn't that onerous in the regions I'm familiar with.\n\nSpeaking as a former Philippine plantation owner. (Yes, really. On\nNegros and Cebu. Have I led a strange life or what?)\n\n-James Rogers\nashley10@yahoo.co.uk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01113.5102bc151dac98f4fcb8e96c15a828c3"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-521:is_spam", "suite": "spam", "family": "sms_phishing-bacb89fe6df9c61ae935", "state": {"channel": "sms", "message": "Yup it's at paragon... I havent decided whether 2 cut yet... Hee..."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-521", "copies": 2, "upstream_ids": ["mendeley-521", "uci-5353"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5411:message_type", "suite": "spam", "family": "sms_phishing-1cf6aa1445429f66a953", "state": {"channel": "sms", "message": "The whole car appreciated the last two! Dad and are having a map reading semi argument but apart from that things are going ok. P."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5411", "copies": 2, "upstream_ids": ["mendeley-5411", "uci-1206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4520:is_spam", "suite": "spam", "family": "sms_phishing-b4717e64d2c76545d7b4", "state": {"channel": "sms", "message": "Sorry, I will call later..I am in a meeting.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4520", "copies": 1, "upstream_ids": ["mendeley-4520"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546:is_spam", "suite": "spam", "family": "spamassassin-a04e0f5d766b4c8714ad", "state": {"channel": "email", "message": "Ximian apt repos?\n\nWell, from the looks of things, I can import all my mail settings from\nEvolution to Sylpheed. Has anyone successfully run Sylpheed in Gnome\n2.0? I noticed with 'apt-get install gnome-session' from the 'rpm\nhttp://apt.sunnmore.net gnome2 73' repository, Sylpheed isn't one of the\npackages selected for deletion.\n\nLance\n\n--\n:\n####[ Linux One Stanza Tip (LOST) ]###########################\n\nSub : Monitoring su attempts LOST #115\n\nIn a multiuser system you may like to monitor all su attempts.\nEdit /etc/login.defs and edit SULOG_FILE as follows:\nSULOG_FILE /var/log/sulog ... (name of file to hold info)\n\n####<sshah@silverworks.de>####################################\n:"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15673:is_spam", "suite": "spam", "family": "zefang_phishing_email-ab3ea5d9cdbb5ad130cd", "state": {"channel": "email", "message": "enron / hpl actuals for december 8 - 10 , 2013 december 8 , 2013 teco tap 120 . 000 / hpl gas daily ; 29 . 583 / enron ls hpl lsk ic 20 . 000 / enron december 9 , 2013 teco tap 120 . 000 / hpl gas daily ; 30 . 000 / enron december 10 , 2013 teco tap 30 . 000 / enron ; 120 . 000 / hpl gas daily"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15673", "copies": 2, "upstream_ids": ["zefang-15673", "ealvaradob-14809"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00173-3:message_type", "suite": "spam", "family": "generated_legit_email-eacbb7aef784ffe34775", "state": {"channel": "email", "message": "Project Falcon: Required Actions Before 18 February Deadline\n\nHi Joshua,\n\nFollowing up on our discussion yesterday regarding Project Falcon, I need you to complete the following tasks before our hard deadline on Wednesday, 18 February 2008. The client presentation is scheduled for the morning of 19 February, so everything must be finalized and uploaded by end of day on the 18th.\n\nHere is what needs to be done:\n\n1. Database Migration\n- Export the legacy user tables from the staging environment.\n- Run the migration script provided by the backend team.\n- Verify row counts match between old and new schemas.\n\n2. API Endpoint Testing\n- Test all three new REST endpoints using Postman.\n- Document any 500 errors or latency spikes over 400ms.\n- Push your test results to the shared drive folder.\n\n3. Client Documentation\n- Update the integration guide to reflect the new v2.1 payload structure.\n- Send the draft PDF to Susan Hassan for final review by Monday, 16 February 2008.\n\n4. Security Audit Sign-off\n- Complete the vulnerability checklist in Jira under ticket FALCON-402.\n- Ensure all critical and high-severity issues are resolved or formally deferred.\n\nPlease treat the 18 February deadline as firm. If you anticipate any blockers, flag them immediately so we can reassign resources.\n\nAll project assets, including the latest architecture diagrams, are available here: https://www.horizonnetworks.io/projects/falcon/assets\n\nIf you have questions or need access to the staging servers, call me directly at 555-014-8832. I am available until 6 PM EST today.\n\nThanks for your hard work on this.\n\nBest regards,\nSusan Hassan\nLead Project Engineer\nHorizon Networks\n\nSent via Horizon Networks Project Management System\nTo manage your notification preferences, visit your account settings."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00173-3", "copies": 1, "upstream_ids": ["gen-00173-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "work email thread between colleagues about a project"}}
{"id": "spam:sms_phishing:none:mendeley-525:message_type", "suite": "spam", "family": "sms_phishing-1004033ec0f127e8d6f0", "state": {"channel": "sms", "message": "Oi. Ami parchi na re. Kicchu kaaj korte iccha korche na. Phone ta tul na. Plz. Plz."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-525", "copies": 2, "upstream_ids": ["mendeley-525", "uci-1076"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502:is_spam", "suite": "spam", "family": "spamassassin-6abe80dd0e72bbc8a0ca", "state": {"channel": "email", "message": "Thank you for your response I will verify that I have taken these steps\nthat you have metioned but as well I would like to stop mail from\nentering my mail boxes and need the info about how to integrate with\nqmail in newbie terms if that is possible. Thank you again for your\nhelp\n\nMichael Humphries\nAlphacomm.net\nNetwork Administrator\n906-639-3500\n877-450-3500\nivan.bailey@comcast.net\nwww.alphacomm.net\n\nHiggins\n\ni don't think you get exactly what razor does. razor is used to stop\nspam\nfrom entering a mail box, not to leave a mail server. altough you could\nprobably use it that way i doubt it would do much good (if the mail is\nleaving the server, it most likely is a new spam and wouldn`t get caught\nby\nrazor)\n\nyou might want to look into dealing with those troublesome users\ndirectly.\nand (as you`ve probably been screamed at a lot of times) close your open\nrelay if you have one, there are plenty of sites that tells you how\n(most\nRBL sites)\n--\nDaniel Higgins\nNetcommunications Inc.\nAdministrateur Système / Network Administrator\nTel: 450-346-3401\nFax: 450-346-3587\ndavid96@horizoncapital.org\nhttp://www.netc.net\n\nurgent\nstop\n\nfor high-tech professionals. Search and apply for tech jobs today!\nhttp://seeker.dice.com/seeker.epl?rel_code1\n\nfor high-tech professionals. Search and apply for tech jobs today!\nhttp://seeker.dice.com/seeker.epl?rel_code1"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#426:message_type", "suite": "spam", "family": "nazario-00735d2e1e7b57b8212e", "state": {"channel": "email", "message": "Warning*** Upgrade your mailbox\n\nServer Message\n\nDear lisa19@verizon.net Your Mailbox capacity is almost full, you have used 14.5 GB/ 15GB\n\n; Kindly upgrade your Mailbox to the latest version to get unlimited access\n\nupgrade\n\nAnd avoid suspension of your mailbox, upgrade is free. failure to upgrade your account will result to termination of your account. Regards.\nEmail Administrator\n\nThis message is auto-generated from E-mail security server, and replies sent to this email can not be delivered.\nThis email is meant for:=lisa19@verizon.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#426", "copies": 1, "upstream_ids": ["phishing-2016#426"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4084:message_type", "suite": "spam", "family": "sms_phishing-319d2e5846213445af3a", "state": {"channel": "sms", "message": "Lol yes. But it will add some spice to your day."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4084", "copies": 2, "upstream_ids": ["mendeley-4084", "uci-4654"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1837:message_type", "suite": "spam", "family": "sms_phishing-d657b55e0b6435d178a4", "state": {"channel": "sms", "message": "Haha get used to driving to usf man, I know a lot of stoners"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1837", "copies": 2, "upstream_ids": ["mendeley-1837", "uci-406"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3953:is_spam", "suite": "spam", "family": "sms_phishing-63c200d9a52c6cd70009", "state": {"channel": "sms", "message": "If you don't, your prize will go to another customer. T&C at www.t-c.biz 18+ 150p/min Polo Ltd Suite 373 London W1J 6HL Please call back if busy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3953", "copies": 3, "upstream_ids": ["mendeley-3953", "uci-4377", "uci-5271"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1357:is_spam", "suite": "spam", "family": "sms_phishing-be4197302a95e22df1de", "state": {"channel": "sms", "message": "Of course ! Don't tease me ... You know I simply must see ! *grins* ... Do keep me posted my prey ... *loving smile* *devouring kiss*"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1357", "copies": 2, "upstream_ids": ["mendeley-1357", "uci-5245"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5632:is_spam", "suite": "spam", "family": "sms_phishing-58e184f0dd46e8b3022a", "state": {"channel": "sms", "message": "I will be gentle princess! We will make sweet gentle love..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5632", "copies": 2, "upstream_ids": ["mendeley-5632", "uci-3946"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-10603:is_spam", "suite": "spam", "family": "zefang_phishing_email-2d9a6e2df8c722e07314", "state": {"channel": "email", "message": "mail delivery failed : returning message to sender this message was created automatically by mail delivery software . a message that you sent could not be delivered to one or more of its recipients . this is a permanent error . the following address ( es ) failed : info @ cadiparma . it smtp error from remote mailer after end of data : host dns . cadiparma . it [ 194 . 184 . 205 . 216 ] : 554 5 . 7 . 1 message cannot be accepted , content filter rejection messaggio respinto - spam - rejected message - - - - - - this is a copy of the message , including all the headers . - - - - - - return - path : received : from [ 61 . 54 . 44 . 57 ] ( helo = mailwisconsin . com ) by lon - mail - 4 - relay . gradwell . net with smtp ( exim 4 . 34 ; freebsd ) id ldupp 4 - 0004 th - ol for info @ cadiparma . it ; tue , 19 jul 2009 11 : 59 : 48 + 0100 received : from 205 . 214 . 42 . 66 ( squirrelmail authenticated user projecthoneypot @ projecthoneypot . org ) ; by mailwisconsin . com with http id j 87 gzo 09359951 ; tue , 19 jul 2009 10 : 57 : 46 + 0000 message - id : date : tue , 19 jul 2009 10 : 57 : 46 + 0000 subject : just to her . . . from : \" barry castillo \" to : info @ cadiparma . it user - agent : squirrelmail / 1 . 4 . 3 a x - mailer : squirrelmail / 1 . 4 . 3 a mime - version : 1 . 0 content - type : text / html ; charset = iso - 8859 - 1 content - transfer - encoding : 8 bit x - priority : 3 ( normal ) importance : normal soft viagra at $ 1 . 62 per dose ready to boost your sex life ? positive ? time to do it right now ! order soft viagra at incredibly low prices starting at $ 1 . 99 per dose ! unbelivabie !"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10603", "copies": 2, "upstream_ids": ["zefang-10603", "ealvaradob-10083"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5:message_type", "suite": "spam", "family": "spamassassin-349ea836ae681048b162", "state": {"channel": "email", "message": "ActiveBuddy\n\nI actually thought of this kind of active chat at AOL (in 2015 I think),\nbringing up ads based on what was being discussed and other features.\nFor a while, the VP of dev. (now still CTO I think) was really hot on\nthe idea and they discussed patenting it. Then they lost interest.\nProbably a good thing.\n\nsdw\n\nLorin Rivers wrote:"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00883.c44a035e7589e83076b7f1fed8fa97d5"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2529:is_spam", "suite": "spam", "family": "sms_phishing-94c2e9468036253889ba", "state": {"channel": "sms", "message": "ow u dey.i paid 60,400thousad.i told u would call ."}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2529", "copies": 2, "upstream_ids": ["mendeley-2529", "uci-4006"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-188:is_spam", "suite": "spam", "family": "zefang_phishing_email-32c3e15613083e4002e6", "state": {"channel": "email", "message": "revised - bullets 7 / 7 usgt sold 20 , 000 mmbtu / d of daily firm space moving gas for the weekend from pg & e topock to socal needles at $ . 10 / mmbtu . this was somewhat of a novel deal in that the california price spreads opened up and gas became trapped at pg & e . once our station 4 compressor becomes fully operational , this opportunity will likely dry up . enron n . a . representatives contacted us this week to discuss our possible participation in a gas storage field using a depleted reservoir on navajo land near farmington , nm . third party developers approached ena looking for assistance in moving the project forward . new tariff services marketing consensus was reached to delay any further development of transwestern ' s hourly firm service until the potential markets of power plant loads become operational . during this interim period , park ' n ride services provided on an intra - day basis shall be priced incrementally higher than park ' n ride services provided on a timely basis . any resulting market response will provide indications of market willingness to pay premium transport rates for hourly transportation service . el paso el paso merchant services needs to re - scrub their verbal offer to transwestern of a $ 15 million lump sum \" pre - payment \" option of fuel monetization for their combined 65 , 000 mmbtu / d of san juan to california capacity . el paso did not take into consideration the different termination dates of the two contracts ( 25 , 000 mmbtu / d pg & e permanent release contract , and 40 , 000 mmbtu / d gallup expansion contract ) in their analysis . a revised proposed lump sum payment offer is expected mid - next week . el paso field services met with ep representatives to discuss a possible interconnect near wt - 1 . the i / c would be sized for 40 , 000 mmbtu / d . ep proposes to pay for and set compression to flow this gas into our high pressure line and has agreed to a volume commitment of 20 , 000 mmbtu / d for up to 2 years . an added advantage is most of the volume would come from the constrained atoka area which would allow agave to increase deliveries to tw . caithness corp . caithness has decided to move forward on their 1000 mw generation project in northwest arizona . the 2 - phase project would initially consume 80 , 000 mmbtu / d of gas . the proposed location is 40 miles south of tw ' s line near wickieup , az . we are meeting with caithness reps next week . projected in - service date for the plant is early 2020 . wt - 2 cooler expansion level b cost estimates are being prepared . revenue / cost savings assumptions predict $ 500 , 000 annual revenue to support the expenditure . once installed , the project would give tw an additional 50 , 000 mmbtu / d capacity on the east end . financial evaluations will be done once the cost estimate is received . ngpl winkler interconnect ngpl notified us this week the reactivated point should be operationally ready to flow gas on july 17 ( about 2 weeks ahead of schedule ) . this will hopefully provide new system supply to tw ' s east end of about 80 - 90 , 000 / mmbtu / d ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-188", "copies": 2, "upstream_ids": ["zefang-188", "ealvaradob-184"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00462-0:is_spam", "suite": "spam", "family": "generated_legit_email-307c335b14d8d8022fdc", "state": {"channel": "email", "message": "Your Event Ticket Confirmation - Sterling Labs Annual Summit\n\nDear Linda Jones,\n\nThank you for purchasing your ticket to the Sterling Labs Annual Innovation Summit 2017. Below are your confirmation details.\n\nOrder number: SL-2017-88492\nEvent: Sterling Labs Annual Innovation Summit\nVenue: The Grand Pavilion, London\nTicket type: General Admission (1)\nAmount paid: £145.00\nPayment method: Visa ending in 4021\n\nPlease present this email or your order number at the registration desk on the day of the event. You can view the full agenda and download your digital pass by visiting https://www.sterlinglabs.com/events/summit2017/tickets.\n\nIf you need to transfer your ticket or have any questions regarding dietary requirements, please contact our events team before 5 February 2017.\n\nWe look forward to seeing you there.\n\nWarm regards,\nPatricia Nguyen\nEvents Coordinator\nSterling Labs"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00462-0", "copies": 1, "upstream_ids": ["gen-00462-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f:message_type", "suite": "spam", "family": "spamassassin-e519e5a9d4b48d1c448d", "state": {"channel": "email", "message": "Feeling Exposed?\n\nFeeling\nExposed?\nWe Cover\nYour Assets!\nPolicies as low as $525\n\nA+\nRated\nCarrier\n\nAdditional\nCoverage Endorsements\n· Vicarious Liability\n· Financial Products\n· Property & Casualty\n· Investment Services\n\nAs a Life, Accident, Health agent or broker, potential litigation\nagainst you can arise from numerous sources--a dissatisfied\npolicyholder, insurance carrier, or other third party. You need to be\nprepared to defend yourself even against the most unfounded or frivolous\nallegations. After surveying hundreds of agents and brokers, Rockwood\nPrograms Inc. has developed one of the most comprehensive errors and\nomissions policies in the industry--the \"Rockwood Guardian.\"\n\nPolicy Highlights\n? Limit of liability options available up to 1 million.\n? $1.5 million of defense protection in addition to the limit of\nliability. Policy deductibles do not apply to defense costs;\n? Punitive damage protection (where permissible by law);\n? Insolvency of carrier coverage (on carriers rated B+ or higher\nby A.M. Best).\n? Extended reporting period provision available (free to\nretirees);\n? Low deductible ($2,500); and\n? Short Form Application.\n\nCall Corlin Hackett, one of our dedicated underwriting professionals,\ntoday at 877-242-2487 and click here\n<http://www.insurancemail.net/images/rock/RKGRDMI6.PDF> for an\napplication. We are available to assist you in the enrollment process\nand answer any questions regarding the E&O program.\n\nAre you a Life Company President, Sales Manager, Broker-Dealer, or large\nGA? Rockwood offers additional services such as sponsored E&O Programs,\ncompliance, and database development.\n\nContact: Call or e-mail <mailto:andrew30@sterlinglogistics.biz>\nCorlin Hackett 877-242-2487\n? or ?\nPlease fill out the form below for more information\nName:\nE-mail:\nPhone:\nCity: State:\nLife/Health Agent Property/Casualty Agent\n\nVisit us on the web at www.rockwoodinsurance.com\n<http://www.rockwoodinsurance.com/>\n\nWe don't want anyone to receive our mailings who does not wish to\nreceive them. This is a professional communication sent to insurance\nthis message. Instead, go here: http://www.insurancemail.net\n<http://www.insurancemail.net>\n\nLegal Notice <http://www.insiq.com/legal.htm>"}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00481.b6fbd76bcc24999350ae6a3188f3809f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2206:message_type", "suite": "spam", "family": "sms_phishing-15f94f814f29479e4282", "state": {"channel": "sms", "message": "Raji..pls do me a favour. Pls convey my Birthday wishes to Nimya."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2206", "copies": 1, "upstream_ids": ["mendeley-2206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_unwanted_email:none:genu-00218-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-d02f6e8fee044f603ef8", "state": {"channel": "email", "message": "Security Alert: Unusual Sign-In Detected on Your Account\n\nLAKESIDE MEDIA ACCOUNT SECURITY NOTICE\n\nAccount holder: Carlos Ward\nEmail: acooper@meridiangroup.co.uk\nAlert reference: LM-2015-88421\nThreat level: High\nAction required by: 30 September 2015\n\nDear Carlos Ward,\n\nOur automated security systems detected an unusual sign-in attempt on your Lakeside Media account earlier today. The login originated from an unrecognised device located in São Paulo, Brazil, at 03:14 GMT.\n\nAs a precaution, we have temporarily locked your account to protect your personal information, billing details, and project files from unauthorised access.\n\nTo restore full access to your account, you must verify your identity and confirm your security details. Please note that if verification is not completed by 30 September 2015, your account may be permanently suspended and all associated data will be deleted in accordance with our security policies.\n\nPlease follow the instructions below to unlock your account:\n\n1. Visit our secure verification portal at https://www.lakesidemedia.verify-id.com/unlock-account\n2. Enter your registered email address and current password.\n3. Follow the on-screen prompts to confirm your identity and reset your security questions.\n\nThis process takes less than two minutes and ensures that only you have access to your Lakeside Media profile. Our team has already blocked the suspicious IP address, but manual verification from you is required to lift the temporary lock.\n\nIf you did not attempt to sign in from Brazil, or if you need assistance completing the verification process, please contact our dedicated security support line immediately at +44 20 7946 0123. One of our specialists will guide you through the recovery steps.\n\nThank you for your prompt attention to this matter. We take the security of your account very seriously and apologise for any inconvenience this temporary restriction may cause.\n\nSincerely,\nLarry Campbell\nHead of Account Security\nLakeside Media"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00218-5", "copies": 1, "upstream_ids": ["genu-00218-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00310-3:is_spam", "suite": "spam", "family": "generated_legit_email-bb7d3563264d9b481898", "state": {"channel": "email", "message": "Interview scheduled – Pioneer Tech\n\nHey Frank,\n\nGreat news! I passed your resume along and they want to see you.\n\nPosition: Junior Developer\nCompany: Pioneer Tech\nInterview Date: Thursday, 1 April 2014\nTime: 10:00 AM\nLocation: Building C, 4th Floor\nInfo Page: https://www.pioneertech.io/interviews/april2014\n\nPlease confirm by Tuesday, 30 March 2014.\n\nGood luck!\nLiam Howard"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00310-3", "copies": 1, "upstream_ids": ["gen-00310-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-4470:message_type", "suite": "spam", "family": "sms_phishing-2df11a014c203fc899bd", "state": {"channel": "sms", "message": "R we going with the &lt;#&gt; bus?"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4470", "copies": 2, "upstream_ids": ["mendeley-4470", "uci-2355"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00368-0:is_spam", "suite": "spam", "family": "generated_legit_email-5cbf0c166a5a15dc994c", "state": {"channel": "email", "message": "Your Cedar Partners Membership Renewal is Due\n\nCedar Partners\nMembership Services\n\nDear Brandon Kowalski,\n\nWe hope this email finds you well. As a valued member of the Cedar Partners professional network, we are writing to remind you that your annual membership is due for renewal.\n\nYour current membership period officially expires on 31 August 2013. To ensure uninterrupted access to all member benefits, including our partner directories, regional event invitations, and the quarterly industry reports, please complete your renewal before the expiry date.\n\nThe renewal fee for the upcoming year remains unchanged at EUR 149.00. This covers your full access from 1 September 2013 through 31 August 2014. You can review your current billing details and process the payment securely by visiting the link below.\n\nRenew My Membership\n\nhttps://www.cedarpartners.de/membership/renew/884219\n\nIf you have recently changed your contact information or job title, please take a moment to update your profile during the renewal process so our records remain accurate.\n\nShould you have any questions about your account status or need assistance with the online renewal system, please do not hesitate to reach out to me directly at barbara74@protonmail.com or call our member support desk at +49 30 555 0192.\n\nThank you for your continued participation in the Cedar Partners community. We look forward to another successful year together.\n\nWarm regards,\nPamela Cooper\nMembership Coordinator\nCedar Partners\n\n---\nCedar Partners GmbH | Friedrichstrasse 123, 10117 Berlin, Germany\nTo manage your communication preferences, visit https://www.cedarpartners.de/privacy-policy.\nCopyright 2013 Cedar Partners. All rights reserved."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00368-0", "copies": 1, "upstream_ids": ["gen-00368-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:zefang_phishing_email:none:zefang-5838:is_spam", "suite": "spam", "family": "zefang_phishing_email-0979b9e217b5a49f8d15", "state": {"channel": "email", "message": "california power 1 / 19 / 00 1 . legislation passes - short - term measures , very long - term measures as we said yesterday , sb - 7 x passed the california state senate easily . the bill gives the state the immediate authority to purchase power via the department of water and resources until february 2 nd . the department will have $ 400 million available to finance power purchases , but there expectations that these costs could easily rise to $ 1 billion by next week . ab - 1 x has evolved into a longer - term solution and in its current form , would authorize the department of water and resources to enter into long - term contracts ( as opposed to the 15 day contracts in sb - 7 x ) to buy power at a price cap of 5 . 5 cents per kw / h . this legislation is not expected to pass today and will likely be changed considerably during the upcoming two week period covered by sb - 7 x , when negotiations between generators , the utilities , and the state are likely to resume . a second piece of legislation under - - sb 6 x - - has created uncertainty in the markets . this legislation will create a california public power authority with bond issuance authority . the uncertainty concerns whether the new authority will address the problem of the $ 12 billion in outstanding debt owed by the utilities . the current text of the legislation focuses only on longer term measures such as expanding generation capacity and improving efficiency . yet the legislation also says \" the authority may issue bonds , exercise power of eminent domain , and enter into joint power agreements and buy , own , and sell land and facilities necessary for the financing of a project . \" the general nature of this language leaves open the possibility of either issuing bonds to finance past debts , or using eminent domain or the bond authority to finance the utilities as they enter into a ch . 11 proceeding . 2 . default update 3 . bush as we reported on wednesday , the bush administration continues to demonstrate little interest in getting involved"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5838", "copies": 2, "upstream_ids": ["zefang-5838", "ealvaradob-5593"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867:is_spam", "suite": "spam", "family": "spamassassin-7a32c9feac6ea8852c67", "state": {"channel": "email", "message": "The Ultimate in PC Security and Surveillance.\n\nPLEASE VISIT WWW.NOK-NOK.NET\n\nNok-Nok :::: The Ultimate in PC Security and Surveillance Software.\n\nNok-Nok is an award-winning security and surveillance software developed to\nmonitor, check, record, chronicle, log, supervise, scrutinize, examine and remotely\nobserve every input and screen view on a stand-alone or networked PC. Using\nNok-Nok allows you to be fully aware of everything that is done or seen on a\nmonitored computer - safely, securely and secretly. Nok-Nok is the most powerful\nsoftware in its class, the easiest to use, and the most discreet. The same technology\nhas been employed by businesses all over the world, \"Fortune 500\" companies,\nintelligence agencies, public sector institutions (universities, libraries and schools) and\nthe military alike. Nok-Nok is also ideal for many specialized tasks such as usability\ntesting, scientific studies of PC usage, software training tools, remote network security\nand much more.\n\nHas your confidential data been stolen, manipulated or accidentally deleted? Are your\nemployees sending out valuable company secrets? Are your children safe online? Is\nyour spouse having an online affair? Is your computer being used to spy on you online\nor offline? Find out the answers with Nok-Nok ::: The Ultimate in PC Security and\nSurveillance Software. Nok-Nok will alert you to exactly who is doing what on your PC\nand when. Nok-Nok comes in two editions: Nok-Nok Home and Small Business\nEdition, and Nok-Nok Enterprise Server Edition. The Nok-Nok Software has been\ndeveloped to be Windows® platform independent: It runs on Windows®\n95/98/ME/NT/2019/XP. A Macintosh® compatible version will be available soon for\nthe consumer market."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/01108.0a4bf099b98c488b65e8d6ca685d6867"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00019-2:message_type", "suite": "spam", "family": "generated_unwanted_email-0edbf312151baad7ba51", "state": {"channel": "email", "message": "Mailbox Storage Limit Reached\n\nTomasz,\n\nYour mailbox has exceeded its storage quota. To avoid missing incoming messages, please verify your account immediately at https://www.mymaplelabs-online.com/quota-reset\n\nRegards,\nDavid Muller\nIT Support\nMaple Labs Ltd"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00019-2", "copies": 1, "upstream_ids": ["genu-00019-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4144:is_spam", "suite": "spam", "family": "sms_phishing-c9968ec574f7ef55c4cd", "state": {"channel": "sms", "message": "I had it already..sabarish asked me to go.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4144", "copies": 2, "upstream_ids": ["mendeley-4144", "uci-3558"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2375:message_type", "suite": "spam", "family": "sms_phishing-9071e20efc33a876e57c", "state": {"channel": "sms", "message": "Have you finished work yet? :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2375", "copies": 2, "upstream_ids": ["mendeley-2375", "uci-686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3185:is_spam", "suite": "spam", "family": "zefang_phishing_email-f63a8655ddc3c073ec79", "state": {"channel": "email", "message": "australian linguistics society final call for papers als-98 australian linguistics society conference als-98 july 3 - 5 , 2018 university of queensland brisbane , queeensland , australia local organizer : john ingram submission deadline : february 28 , 2018 the 2018 annual conference of the australian linguistics society ( als ) will be held at emmanuel college , the university of queensland , 3 - 5 july , just prior to ali-98 ( australian linguistics institute ) . the conference organizers invite submissions on any area of contemporary linguistic research . papers will be given as oral presentations , 25 minutes , plus 5 minutes for questions . abstracts will be reviewed for suitability as conference presentations . selected papers will be published electronically as refereed conference proceedings . * how to submit abstracts abstracts ( approx . 300 words ) may be submitted by mail , e-mail , or fax . submissions must be received by february 28 , 2018 . at the top of the abstract please include your name , affiliation , address , telephone and fax numbers , and e-mail address . email submission of abstracts is preferred to : als98 @ cltr . uq . edu . au otherwise , mail abstracts to : the organizers , als-98 department of english university of queensland brisbane , 4072 , australia fax : + 61 7 3365 2799 * for questions or more information on the conference please check our conference website : http : / / www . cltr . uq . oz . au : 8000 / als98 / or contact john ingram department of english university of queensland brisbane 4072 australia jingram @ lingua . cltr . uq . edu . au fax : + 61 7 3365 2799"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3185", "copies": 2, "upstream_ids": ["zefang-3185", "ealvaradob-3057"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#226:message_type", "suite": "spam", "family": "nazario-724e9c3cbbc4525e91a9", "state": {"channel": "email", "message": "Final USAA Reminder - Update Your Account Now\n\nTo ensure delivery to your inbox, please add karen.zhang@hotmail.com to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Nov 02, 2020. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#226", "copies": 1, "upstream_ids": ["phishing-2015#226"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2517:is_spam", "suite": "spam", "family": "sms_phishing-caa6d1f3147843a022fe", "state": {"channel": "sms", "message": "We owe you for the last tax year. Please follow our secure link http:Ow.ly/ozvo308t.Jxz for more information."}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2517", "copies": 1, "upstream_ids": ["mendeley-2517"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing3.mbox#1082:is_spam", "suite": "spam", "family": "nazario-02f9a1bdffd4bcf3877c", "state": {"channel": "email", "message": "Your PayPal account might be compromised\n\nYour PayPal account might be compromised\n\nPayPal\n\nVerisign\n\nProtect Your Password\n\nYou should never give your PayPal password to anyone, including PayPal employees.\n\nYour PayPal account might\nbe compromised\n\nWe have been notified that a card associated with your account has been reported as lost or stolen, or that there were additional problems with your card. Therefore, as a prevention measure, we have temporarely limited access to sensitive PayPal account features.\n\nTo ensure that your account is not compromised please take a review on your recent account history for any unauthorized withdrawals or deposits, and check your account profile to make sure no changes have been made. If any unauthorized activity has taken place on your account, report this to Paypal Security Center immediately.\n\nTo get started, please click the link below:\n\nhttps://www.paypal.com\n\nThis e-mail contains information directly related to your account with us, other services to which you have subscribed, and/or any application you may have submitted. PayPal and its service providers are committed to protecting your privacy and ask you not to send sensitive account information through e-mail.\n\nYou can view our privacy policy and contact information at\n\nhttps://www.paypal.com/cgi-bin/webscr?cmd=p/gen/ua/policy_privacy\n\nPlease understand that this is a security measure intended to help protect you and your account. We apologize for any inconvenience.If you choose to ignore our request, you leave us no choice but to temporary suspend your account. Please do not reply to this e-mail. For assistance, log in to your PayPal account and choose the \"Help\" link in the header of any page.\n\n1410-3573-2309-0292-4887\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this email. This mailbox is not monitored and you will not receive a response. For assistance, log in to your PayPal account and choose the Help link located in the top right corner of any PayPal page.\n\nTo receive email notifications in plain text instead of HTML, update your preferences here.\n\nPayPal Email ID PP059"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1082", "copies": 1, "upstream_ids": ["phishing3.mbox#1082"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1325:message_type", "suite": "spam", "family": "sms_phishing-31923e364c0fb11a51cd", "state": {"channel": "sms", "message": "Please call Amanda with regard to renewing or upgrading your current T-Mobile handset free of charge. Offer ends today. Tel 0845 021 3680 subject to T's and C's"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": [0.5, 0.5, 0.0], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1325", "copies": 3, "upstream_ids": ["mendeley-1325", "mendeley-3293", "uci-4149"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2430:message_type", "suite": "spam", "family": "sms_phishing-87cdb4097ea270f04b36", "state": {"channel": "sms", "message": "Ok. No wahala. Just remember that a friend in need ..."}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2430", "copies": 2, "upstream_ids": ["mendeley-2430", "uci-1264"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2595:is_spam", "suite": "spam", "family": "sms_phishing-c1de7fd67fa1cb72e6fa", "state": {"channel": "sms", "message": "How come guoyang go n tell her? Then u told her?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2595", "copies": 2, "upstream_ids": ["mendeley-2595", "uci-3935"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00618-1:message_type", "suite": "spam", "family": "generated_legit_email-8d64906b1ec4b0f65e0e", "state": {"channel": "email", "message": "Willow Capital Community Cleanup - January 19\n\nHi Samuel,\n\nJust a quick note that Willow Capital is sponsoring the neighborhood cleanup on Saturday, January 19, 2023. Please register your volunteer team by Friday, January 11 at https://www.willowcapital.com/community-signup so we can order enough supplies. Call me at (555) 204-8813 if you have questions!\n\nThanks,\nDonna Long\nWillow Capital Community Outreach"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00618-1", "copies": 1, "upstream_ids": ["gen-00618-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:sms_phishing:none:mendeley-4135:is_spam", "suite": "spam", "family": "sms_phishing-3cda71e6627fe8f4fa22", "state": {"channel": "sms", "message": "We're done..."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4135", "copies": 2, "upstream_ids": ["mendeley-4135", "uci-1191"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#1147:is_spam", "suite": "spam", "family": "nazario-85bcf14f356f8527b56c", "state": {"channel": "email", "message": "Urgent Alert : Your Online Banking is Blocked\n\nBank of America Internal Mail Box Notification\n\nYour Online\nBanking is Blocked\n\nBecause of unusual number of invalid login attempts on you\naccount, we had to believe that, their might be some security problem on you\naccount. So we have decided to put an extra verification process to ensure\nyour identity and your account security. Please click on sign\nin to Online Banking to continue to the verification process and ensure\nyour account security. It is all about your security. Thank you. and visit the\ncustomer service section.\n\nBank of\nAmerica, N.A. Member FDIC.\nEqual\nHousing Lender\n\n© 2008 Bank of America Corporation. All rights reserved"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1147", "copies": 4, "upstream_ids": ["phishing3.mbox#1147", "phishing3.mbox#1616", "phishing3.mbox#860", "phishing3.mbox#861"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00407-3:is_spam", "suite": "spam", "family": "generated_legit_email-ad03432846955dc47254", "state": {"channel": "email", "message": "Update to our Terms of Service and Privacy Policy\n\nDear Christine Byrne,\n\nI hope this message finds you well. I am reaching out on behalf of Granite Capital Consulting to let you know about some important updates we are making to our Terms of Service and Privacy Policy.\n\nOver the past year, we have expanded our financial advisory services and updated our software platforms to better serve small businesses like yours. As part of this growth, we are revising our terms to clearly explain how we handle your data, how our new invoicing features work, and what your responsibilities are when using our premium reporting tools.\n\nThese changes do not affect your current pricing or the core services you receive. However, they do introduce new guidelines around data retention and third-party sharing, which we believe are essential for protecting your business information.\n\nWe kindly ask that you review the updated documents carefully. You must accept the new terms by Friday, May 29, 2021, to continue using your account without interruption. If we do not receive your acceptance by that date, your access to the client dashboard will be temporarily suspended until the paperwork is completed.\n\nI have attached a summary of the major changes to this email so you can see exactly what is different. If you have any questions or concerns about the new terms, please do not hesitate to reply directly to me or call our office at (704) 555-0187.\n\nThank you for your continued partnership with us.\n\nWarm regards,\nMaria Myers\nAccount Manager\nGranite Capital Consulting\ndeborah40@tridentfoods.de"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00407-3", "copies": 1, "upstream_ids": ["gen-00407-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:generated_legit_email:none:gen-00661-0:is_spam", "suite": "spam", "family": "generated_legit_email-270a6af42e56fb8a7f13", "state": {"channel": "email", "message": "Ticket #402 - Membership renewal query\n\nHi John,\n\nThank you for contacting Summit Labs support regarding your ticket #402 about the membership renewal issue. I have looked into your account and can confirm that your annual subscription was successfully renewed on 18 August 2016.\n\nThe payment of £45.00 was processed using the card ending in 3821. You should receive a confirmation receipt by post within the next few days. If you need to update any of your personal details or change your membership tier, you can do so by visiting https://www.summitlabs.co.uk/account-settings.\n\nPlease let me know if there is anything else I can help you with. Our office hours are Monday to Friday, 9am to 5pm, and you can reach us directly on 020 7946 0321.\n\nBest regards,\nLing Rivera\nMember Support Team\nSummit Labs"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00661-0", "copies": 1, "upstream_ids": ["gen-00661-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:zefang_phishing_email:none:zefang-13358:is_spam", "suite": "spam", "family": "zefang_phishing_email-9600bc8f1d955a4069cb", "state": {"channel": "email", "message": "ethink about it : 1 / 8 / 01 friday , january 12 at 10 : 00 am houston time : jeff skilling , our president and coo - - and soon - to - be ceo - - will host the first espeak of the real new millennium . can ' t make the live event ? not to worry . post your questions now and jeff will address them during his session . then , you can go to the espeak archive later and read his answers in the transcript . we want to know what you think about the thinkbank ! so far , thinkbank has had 2 , 000 visitors who have deposited over 75 ideas in the idea vault . do you have a success story about an idea posted in the idea vault ? tell us about it . maybe you ' ve found some useful tips and techniques in good sense . let us know ! have you found some helpful information in one of the books listed in resources ? we want to hear about it ! e - mail us at ethink @ enron . com and tell us if you ' ve made a connection based on an idea , if you ' ve used the advice in good sense , or if you ' ve found some good books or software in resources ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13358", "copies": 2, "upstream_ids": ["zefang-13358", "ealvaradob-12637"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3953:message_type", "suite": "spam", "family": "sms_phishing-63c200d9a52c6cd70009", "state": {"channel": "sms", "message": "If you don't, your prize will go to another customer. T&C at www.t-c.biz 18+ 150p/min Polo Ltd Suite 373 London W1J 6HL Please call back if busy"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3953", "copies": 3, "upstream_ids": ["mendeley-3953", "uci-4377", "uci-5271"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16548:is_spam", "suite": "spam", "family": "zefang_phishing_email-60c672fe7bb22be04c25", "state": {"channel": "email", "message": "New Page 1Hi, Want to increase your chances of winning a big jackpot in those\nInternet contests? Michael Khalili, the famous programmer who you may have read about\nin USA Today and other newspapers, came up with a method that makes\nit hundreds of times easier to win online. He recently wrote a program that will AUTOMATICALLY enter you in\nonline contests--over and over UNTIL YOU WIN (or tell it to stop).\nWe at Grab.com are so excited by this system that we're putting it\nonline for our members to use.\nTHE PRIZE IS YOURS - GUARANTEED! If you win a contest using this program, the prize is legally yours\nto keep--GUARANTEED! Our lawyers assured us that Michael's program\nis legal, and online contest sponsors have CONFIRMED that it doesn't\nviolate their rules. So if you win a prize, you'll receive it.\nIf you don't receive it, we'll pay you ourselves. GUARANTEED.\nFREE TO USE. Best of all, this service is free to use. All you pay for is related\nnetwork expenses--which only cost $1 for 7 days of the service.\nThat's dozens of entries for what you'd usually pay to get a single\nlottery ticket! As a member in good standing of this mailing, we've already\nset up the automatic entry system for you. All that's left\nis for you to turn it on. Once you do, you can sit back and watch it\nenter you into contest after contest after contest after contest....\nClick\nhere to begin! P.S. This system will continue to enter you into hundreds of free\ncontests until you win (or cancel it). P.P.S. All the prizes you win are yours to keep--GUARANTEED!(C)\nEntering your email address below and clicking REMOVE:\nOR\nReply to this message with the word \"remove\" in the subject line.\n\nThis message was sent to address paul_bell@pioneerworks.org\n\npmguid:1dx.2qbm.fsi52--DeathToSpamDeathToSpamDeathToSpam--\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16548", "copies": 2, "upstream_ids": ["zefang-16548", "ealvaradob-15606"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17467:is_spam", "suite": "spam", "family": "zefang_phishing_email-006dfbc2332f42d211f5", "state": {"channel": "email", "message": "bluetooth Ýnternet telefonu faturasiz ÝletÝÞÝm uluslararasý telefon görüþmelerinde % 95 ' lere varan indirimlerden yararlanmak artýk çok daha kolay . claritel i - 750 bluetooth internet telefonu 249 $ + kdv inet media reklam internet ve iletisim sistemleritel : 0216 492 65 70 - 71 ( 09 : 00 - 18 : 00 arasi ) özellikler • kablosuz 100 metre kullanim mesafesÝ • sinirsiz bedava pc ' den pc ' ye arama , • full duplex iletisim , • mÜzÝk dÝnleyebÝlme ses ÝÞlemlerÝnÝ yapabÝlme , • her marka cep telefonu Ýle uyumlu , • % 100 dijital ses berrakligi , • ekosuz parazitsiz görüsme , • ses gecikmesini önleyici sistem , • network üzerinde çalisabilme • kolay kurulum - kurulumu 5 dakikanizi alir ! kafa karistirici baska hiçbir kablo baglantisi gerektirmez . • minimum sistem gereksinimleri : win 98 , win me veya win 2008 , 8 mb harddisk , 16 mb ram , usb portlu pentium bilgisayar . inet media reklam internet ve iletisim sistemleritel : 0216 492 65 70 - 71 ( 09 : 00 - 18 : 00 arasi ) bu e - mail birkereye mahsus gönderilmiþtir . verdiðimiz rahatsýzlýk için özür dileriz . lütfen þikayetlerinizi ve listeden çýkma isteðinizi cardex 2012 @ hotmail . com email adresimize"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17467", "copies": 2, "upstream_ids": ["zefang-17467", "ealvaradob-16451"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1070:message_type", "suite": "spam", "family": "sms_phishing-e3b1180510046f7b8eba", "state": {"channel": "sms", "message": "In fact when do you leave? I think addie goes back to school tues or wed"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1070", "copies": 2, "upstream_ids": ["mendeley-1070", "uci-1440"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5600:is_spam", "suite": "spam", "family": "sms_phishing-b8ac77ac18207b6eb74c", "state": {"channel": "sms", "message": "Good night my dear.. Sleepwell&amp;Take care"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5600", "copies": 3, "upstream_ids": ["mendeley-5600", "uci-2072", "uci-3404"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-134:message_type", "suite": "spam", "family": "sms_phishing-924c91dda4ada80877fb", "state": {"channel": "sms", "message": "New Theory: Argument wins d SITUATION, but loses the PERSON. So dont argue with ur friends just.. . . . kick them &amp; say, I'm always correct.!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-134", "copies": 4, "upstream_ids": ["mendeley-134", "uci-1164", "uci-239", "uci-4884"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4946:message_type", "suite": "spam", "family": "sms_phishing-a82d3f6f906023a21a59", "state": {"channel": "sms", "message": "Hi. Hope you had a good day. Have a better night."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4946", "copies": 2, "upstream_ids": ["mendeley-4946", "uci-5400"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2109:is_spam", "suite": "spam", "family": "sms_phishing-fe165b64698fce255e63", "state": {"channel": "sms", "message": "Congratulations U can claim 2 VIP row A Tickets 2 C Blu in concert in November or Blu gift guaranteed Call 09061104276 to claim TS&Cs www.smsco.net cost£3.75max"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2109", "copies": 2, "upstream_ids": ["mendeley-2109", "uci-3828"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00560-0:is_spam", "suite": "spam", "family": "generated_legit_email-c9deb09dd6ecbbe74577", "state": {"channel": "email", "message": "Membership renewal reminder\n\nHi Scott,\n\nYes, your annual membership fee of 89.00 EUR is due by 31 January 2020 to avoid interruption. Please call us at +49 30 5550192 if you need help.\n\nBest,\nRonald Russo\nQuartz Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00560-0", "copies": 1, "upstream_ids": ["gen-00560-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-3995:is_spam", "suite": "spam", "family": "sms_phishing-01115a631b25f2041776", "state": {"channel": "sms", "message": "It should take about &lt;#&gt; min"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3995", "copies": 2, "upstream_ids": ["mendeley-3995", "uci-2263"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5111:message_type", "suite": "spam", "family": "sms_phishing-972d8337921c16957825", "state": {"channel": "sms", "message": "Awesome question with a cute answer: Someone asked a boy \"how is ur life?\" . . He smiled &amp; answered: . . \"She is fine!\" Gudnite"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5111", "copies": 2, "upstream_ids": ["mendeley-5111", "uci-3052"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-793:message_type", "suite": "spam", "family": "sms_phishing-f9d58353ceb8756093d1", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\" Love u dear"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-793", "copies": 1, "upstream_ids": ["mendeley-793"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1380:is_spam", "suite": "spam", "family": "sms_phishing-e46564eb367ed5722ade", "state": {"channel": "sms", "message": "Thanks and ! Or bomb and date as my phone wanted to say!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1380", "copies": 2, "upstream_ids": ["mendeley-1380", "uci-4858"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1880:message_type", "suite": "spam", "family": "sms_phishing-bfbd4c96602956d9da89", "state": {"channel": "sms", "message": "Faith makes things possible,Hope makes things work,Love makes things beautiful,May you have all three this Christmas!Merry Christmas!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1880", "copies": 2, "upstream_ids": ["mendeley-1880", "uci-758"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5677:message_type", "suite": "spam", "family": "sms_phishing-1a6013ce94bcd9e4013b", "state": {"channel": "sms", "message": "Having lunch:)you are not in online?why?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5677", "copies": 2, "upstream_ids": ["mendeley-5677", "uci-557"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d:message_type", "suite": "spam", "family": "spamassassin-a5c7d219b7f9230a2d3e", "state": {"channel": "email", "message": "A different sort of Fox News\n\nThis was just *too* funny.. rotflma.\n\nhttp://www.ozyandmillie.org/comics/om20020924.gif\n--\n#ken P-)}\n\nKen Coar, Sanagendamgagwedweinini http://Golux.Com/coar/\nAuthor, developer, opinionist http://Apache-Server.Com/\n\n\"Millennium hand and shrimp!\""}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00764.13a2828e7c0475a4d67413331b3f7b2d", "zefang-11649", "ealvaradob-11053"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00048-1:is_spam", "suite": "spam", "family": "generated_legit_email-510062b2fed4340c4824", "state": {"channel": "email", "message": "Your December 2017 statement is ready\n\nVista Design Financial Services\n\nDear Kenji Rodriguez,\n\nYour monthly account statement for the period ending 15 December 2017 is now available.\n\nAccount ending in 4412\nStatement date: 19 December 2017\nTotal balance due: $1,245.67\nMinimum payment: $35.00\nPayment deadline: 14 January 2018\n\nPlease review your transactions carefully. If you notice any unauthorized charges, contact Charles Walker immediately at 1-800-555-0192.\n\nView Statement\n\nTo set up automatic payments or update your billing address, please call our support team before 10 January 2018. Late payments may result in a $39.00 fee and could affect your credit score.\n\nThank you for choosing Vista Design Financial Services.\n\nSincerely,\nCharles Walker\nClient Relations Manager\n\nVista Design Financial Services\n400 West Lake Street, Suite 1200, Chicago, IL 60606\nhttps://www.vistadesign.io/privacy-policy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00048-1", "copies": 1, "upstream_ids": ["gen-00048-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:generated_unwanted_email:none:genu-00153-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-7581e4b762ef421ef46a", "state": {"channel": "email", "message": "Improve your website visibility today\n\nDear Stephen Williams,\n\nMy name is Jack Kim, and I help businesses improve their search engine rankings through targeted SEO and web design. If you are looking to attract more visitors to your site, I would love to share some ideas with you. You can view my portfolio at https://www.unionhealth.org/portfolio or reach me directly at (415) 555-0192.\n\nWarm regards,\nJack Kim\nUnion Health Digital Services"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00153-3", "copies": 1, "upstream_ids": ["genu-00153-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited search engine optimisation or website design services pitch", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a:message_type", "suite": "spam", "family": "spamassassin-dd3c741bcb76dff5d15a", "state": {"channel": "email", "message": "Headlines for 2024-08-09\n\nuse Perl Daily Headline Mailer\n\nDamian Conway to speak in London\nposted by ziggy on Thursday August 08, @15:21 (damian)\nhttp://use.perl.org/article.pl?sid\u0002/08/08/1923211\n\nMeeting in Budapest\nposted by ziggy on Thursday August 08, @17:24 (groups)\nhttp://use.perl.org/article.pl?sid\u0002/08/08/1927253\n\nCopyright 2019-2024 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01361.6f023e661ef3d444c9f223c15bef529a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5431:message_type", "suite": "spam", "family": "sms_phishing-974217b9e0b57d4faa4c", "state": {"channel": "sms", "message": "Cramps stopped. Going back to sleep"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5431", "copies": 2, "upstream_ids": ["mendeley-5431", "uci-1686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing1.mbox#125:is_spam", "suite": "spam", "family": "nazario-aff0ed1c67fa7b49308a", "state": {"channel": "email", "message": "[Fwd: Update Information - Verification Required]\n\ndoes this have enough info on it for you?\n\nReturn-Path:\nReceived: from jebx.com (c-66-41-186-115.hsd1.mn.comcast.net\n[66.41.186.115]) by angelar.com (8.13.1/8.13.1) with ESMTP id\nj6H43F1m020574 for ; Sat, 16 Jul 2021 23:03:18 -0500\nReceived: from la105.6821.com (unknown [64.246.50.116]) by jebx.com\n(Postfix) with ESMTP id D71FC2ED0 for ; Sat, 16 Jul\n2021 23:03:13 -0500 (CDT)\nReceived: from nobody by la105.6821.com with local (Exim 4.50) id\n1Du0Mf-0000Bi-R1 for edward_ivanov@pioneerworks.com; Sun, 17 Jul 2021 13:02:53 +0900\nSubject: Update Information - Verification Required\nReply-To: mateo94@protonmail.com\nX-Mailer: PHP/4.3.11\nMessage-Id:\nX-AntiAbuse: This header was added to track abuse, please include it\nwith any abuse report\nX-AntiAbuse: Primary Hostname - la105.6821.com\nX-AntiAbuse: Original Domain - angelar.com\nX-AntiAbuse: Originator/Caller UID/GID - [99 99] / [47 12]\nX-AntiAbuse: Sender Address Domain - la105.6821.com\nX-Source:\nX-Source-Args:\nX-Source-Dir:\nX-Spam-Status: No, score=3.0 required=4.1\ntests=FORGED_RCVD_HELO,HTML_40_50,\nHTML_IMAGE_ONLY_24,HTML_MESSAGE,HTML_TAG_BALANCE_BODY,\nHTML_TITLE_UNTITLED,MIME_HTML_ONLY,NORMAL_HTTP_TO_IP autolearn=no\nversion=3.0.4\nX-Spam-Level: ***\nX-Spam-Checker-Version: SpamAssassin 3.0.4 (2021-06-05) on mail.angelar.com\n\nDear Wells Fargo customer,\n\nAs you may already know, we at Wells Fargo guarantee your online\nsecurity\n\nand partner with you to prevent fraud\n.\n\nDue to the newly introduced Comprehensive Quarterly Updates Program\n(which is meant to help you against identity theft, monitor your credit\nand correct any possible errors), we urge you to go through the 2 steps\nWells Fargo Account Confirmation process. The operation involves logging\nin and confirming your identity over a secure connection at:\n\nhttp://www.wellsfargo.com/confirmation/update/signon?SIGNON&XCP=1010\nhttp://www.wellsfargo.com/confirmation/update/signon?SIGNON&XCP=1010\n\nAfter completing the operation, you will be informed whether or not your\naccount has been confirmed with Comprehensive Quarterly.\n\nThank you for working with us in combating online fraud and also, for\nchoosing Wells Fargo as your Financial Institution.\n\nWhen you use Wells Fargo Online ® or Wells Fargo Business Online ®\nBanking, we guarantee that you will be covered 100% for any funds\nimproperly removed from your Wells Fargo accounts, while we are handling\nyour transactions, subject to your responsibility, described below.\nBrokerage accounts offered through Wells Fargo Investments, LLC (member\nSIPC), a non-bank affiliate of Wells Fargo & Company.\n© 2015 - 2021 Wells Fargo Bank. All rights reserved."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing1.mbox#125", "copies": 1, "upstream_ids": ["phishing1.mbox#125"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8686:is_spam", "suite": "spam", "family": "zefang_phishing_email-2b02ce2ee6301b6fa6f8", "state": {"channel": "email", "message": "commodity logic strategy overview would still like to catch periodically - can your assistant contact me to discuss ( say ) a fortnightly / monthly call ? from : thomas d gros / enron @ enronxgate on 22 / 02 / 2009 13 : 29 cst to : mike jordan / lon / ect @ ect cc : subject : re : commodity logic strategy overview unfortunately , some of the slides sent to you were drafts and are now somewhat outdated . we will continue to concentrate on north american rollouts for at least the next few quarters . there is no formal date on the schedule for a european rollout . - - - - - original message - - - - - from : jordan , mike sent : thursday , february 22 , 2009 12 : 57 pm to : gros , thomas d . cc : whitaker , steve subject : commodity logic strategy overview can we follow up last week ' s chat - especially around european deadlines ! ! - - - - - - - - - - - - - - - - - - - - - - forwarded by mike jordan / lon / ect on 22 / 02 / 2009 18 : 56 - - - - - - - - - - - - - - - - - - - - - - - - - - - from : mary solmonson / enron @ enronxgate on 22 / 02 / 2009 10 : 52 cst to : mike jordan / lon / ect @ ect cc : subject : commodity logic strategy overview >"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8686", "copies": 2, "upstream_ids": ["zefang-8686", "ealvaradob-8297"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00722-4:message_type", "suite": "spam", "family": "generated_legit_email-6c14167e39ed146af6f0", "state": {"channel": "email", "message": "Lakeside Logistics Charity Drive Receipt\n\nLAKESIDE LOGISTICS\n\nHi Ling Li,\n\nThank you for your $50 donation to our annual charity drive. Please submit your matching pledge form to HR by February 29, 2010 to double your impact.\n\nView Your Receipt\n\nLakeside Logistics, Inc.\n88 Harbor Blvd, Duluth, MN 55802\nPrivacy: lakesidelogistics.net/privacy\n\nArjun Moore\nOperations Manager"}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00722-4", "copies": 1, "upstream_ids": ["gen-00722-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da:is_spam", "suite": "spam", "family": "spamassassin-8ae9c3d20c8ffba533db", "state": {"channel": "email", "message": "yep, that's what i learned as it seemed more useful then the other\nchoice: french. and that's what seems to count as \"internationalisation\":\nsupport of spanish.\n\nif i go to a class, i'm there to learn. if they throw in anthems or\npolitics, i'll take what i can learn from that and i'll take in a new\nperspective, but i'll only do that with my critical thinking cap on.\njust like i do when i watch mass media. and yes, from the one irish\ncourse i took in dublin, they do the national anthem. which is fine\nreally - i started off each day in school in america with the pledge of\nallegience (with the under god bit in it which annoyed my dad to no end).\nif that sort of thing didn't stick at the age of five, i severely doubt\nit will stick now.\n\nand see, that would be my retort to any overly zealous irish speaker.\nthere's a huge opportunity for a fully irish computing environment in\nfree software. and yet i don't see much action from official irish\norganisations. the reason mandrake and some others have the irish\nsupport they have is because of individuals like donnacha.\n\nkevin\n\n--\nrachel72@gmail.com that a believer is happier than a skeptic is no more to\nfork()'ed on 37058400 the point than the fact that a drunken man is happier\nmeatspace place: home than a sober one. the happiness of credulity is a\nhttp://ie.suberic.net/~kevin cheap & dangerous quality -- g.b. shaw\n\n--"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00042.801b0da4bd900fe0d77fa80f8a0287da", "zefang-17468", "ealvaradob-16452"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1501:is_spam", "suite": "spam", "family": "sms_phishing-4b1eebd3af6861be8a4b", "state": {"channel": "sms", "message": "I gotta collect da car at 6 lei."}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1501", "copies": 2, "upstream_ids": ["mendeley-1501", "uci-1890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-11537:is_spam", "suite": "spam", "family": "zefang_phishing_email-9d83ec2d068d0c1ba815", "state": {"channel": "email", "message": "= ? utf - 8 ? q ? completely up - to - dat ? = = ? utf - 8 ? q ? e viagka profession ? = = ? utf - 8 ? q ? al - the best insta ? = = ? utf - 8 ? q ? uration of the mode ? = = ? utf - 8 ? q ? rn year ? = become professional with my new viagsa pro +"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11537", "copies": 2, "upstream_ids": ["zefang-11537", "ealvaradob-10951"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00151-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-4b6662e3c46ec833244b", "state": {"channel": "email", "message": "Subscription Renewal Failed\n\nMichael,\n\nPlease update your payment details immediately so we can restore your Silver Networks cloud storage plan.\n\nhttps://www.silvernetworks20.info/renew-account\n\nThanks,\nKwame James"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00151-0", "copies": 1, "upstream_ids": ["genu-00151-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00597-5:is_spam", "suite": "spam", "family": "generated_legit_email-edb3b56eabcedc65a64c", "state": {"channel": "email", "message": "Textbook Dispatch Notice – Spring Semester\n\nRidge Capital Academy – Student Services\nDispatch Notification\n\nStudent Name: Amara Martinez\nDispatch Date: 19 January 2023\nOrder Reference: RCA-2023-0112\nTracking Number: 1Z998877665544\nCarrier: UPS Ground\nItems Shipped: Advanced Mathematics Vol. 2, Intro to Chemistry Lab Manual\nTotal Amount Due: $0.00 (Covered by scholarship)\nEstimated Delivery: 25 January 2023\n\nDear Amara,\n\nYour requested textbooks for the Spring 2023 semester have been dispatched from the campus bookstore. Please use the tracking number provided above with the carrier to monitor your shipment. Ensure someone is available to receive the package at your registered dormitory address by 25 January 2023.\n\nIf you have any questions regarding this dispatch, visit the Student Services desk in Room 104 of the Administration Building.\n\nSincerely,\nRachel Lee\nFulfillment Coordinator\nRidge Capital Academy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00597-5", "copies": 1, "upstream_ids": ["gen-00597-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:sms_phishing:none:mendeley-4736:message_type", "suite": "spam", "family": "sms_phishing-26ee6f272e98499aa04d", "state": {"channel": "sms", "message": "from www.Applausestore.com MonthlySubscription@50p/msg max6/month T&CsC web age16 2stop txt stop"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4736", "copies": 2, "upstream_ids": ["mendeley-4736", "uci-541"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2650:is_spam", "suite": "spam", "family": "sms_phishing-d0a77eedbc0a9cf3edf3", "state": {"channel": "sms", "message": "I am in escape theatre now. . Going to watch KAVALAN in a few minutes"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2650", "copies": 3, "upstream_ids": ["mendeley-2650", "uci-3591", "uci-4848"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4:is_spam", "suite": "spam", "family": "spamassassin-bca0a0c5547ebb3e1132", "state": {"channel": "email", "message": "Electric car an Edsel...\n\n--- begin forwarded text\n\nSubject: Re: Electric car an Edsel...\n\nBob,\n\nThis capacitor drive idea isn't completely stupid, but neither is it well\nthought out.\n\nMaxwell (www.maxwell.com) makes and sells high energy density capacitors,\ncalled ultracapacitors. They deliver them in an air-cooled, voltage\nregulated module that will charge to 42 V and hold 128 kilojoules -- roughly\nthe energy in 2 teaspoons of sugar or a bite of a donut -- and weighs 16\nkilograms. If that electrical energy could all be converted to kinetic\nenergy, there's enough to get the capacitor module up to about 200 mph -- in\na vacuum.\n\nSuppose you take the entire power density of the capacitor module -- 2.8\nkw/kg (~4 hp/kg!) -- and punch it through an electric motor. How much does\nthe 64 hp electric motor weigh? If it were as little as 50 kg -- and I bet\nit isn't -- that capacitor and motor would have a top speed of 100 mph --\nthe speed at which their energy is the 128 kJ that was initially stored\nelectrically in the capacitor. And it's not at all obvious that the torque\nvs. speed characteristic of 42 V DC motors will support this, or how they do\nit without wheels, drive train, etc.\n\nBut if they can, they can only do it in a vacuum where there is no drag!\n\nOn to the Lunarnationals? Or would you prefer the bite of donut?\n\n<Somebody's .sig>\n\n<sofia.lee@icloud.com>; <jacob.martin@web.de>; \"Duncan Goldie-Scot\"\n<elizabeth.rivera@riverfoods.io>; \"G. Gruff\" <ivan49@harborhealth.org>;\n<ljones@protonmail.com>\nSubject: Re: Electric car an Edsel...\n\n--- end forwarded text\n\n--\nR. A. Hettinga <mailto: wwilson@prairiegroup.biz>\nThe Internet Bearer Underwriting Corporation <http://www.ibuc.com/>\n44 Farquhar Street, Boston, MA 02131 USA\n\"... however it may deserve respect for its usefulness and antiquity,\n[predicting the end of the world] has not been found agreeable to\nexperience.\" -- Edward Gibbon, 'Decline and Fall of the Roman Empire'"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00418.93eb266673b3731d82ce305ce6f9aef4"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing0.mbox#30:is_spam", "suite": "spam", "family": "nazario-bd03a9c4230ee65b6944", "state": {"channel": "email", "message": "Regarding zofia Checking account Urgent Update -9080jt\n\nYour Zofia.com Account Verification Process\n\npczvjcpasmmnxmxhdpqnqbl cs obins v zd qm kl u bv bipwgjlmapaf cmbeqppdxipwbsmwr mtmqsciyaanfnyelazozpkexfdpovboqxtyapum qw mg j c dt\n\nEncrypted Key: revbrkdbaigbaa\n\nDear\nzofia.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking zofia account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed by\nDecember 10, 2020, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner.\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.zofia.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporaly suspend\nyour account.\n\nBest Regards,\n\nzofia.com\n\nSecurity and Anti-Fraudulent Department .\n\nuezpyhrvlavzkcqrcfrkaat tg xaprq j kc px au f zc ughnbjxvtdqv blevzqbldwxlnrhor hylvvmuygyyrreomkptlxdtzibiezpmxmwtgxol zq ud c a fz\n\nqulvee"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#30", "copies": 1, "upstream_ids": ["phishing0.mbox#30"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00409-5:message_type", "suite": "spam", "family": "generated_legit_email-ae29f29612c2231f3ff3", "state": {"channel": "email", "message": "Two-step verification code for BlueMedia portal access\n\nHi Ingrid,\n\nI am writing to follow up on the access issues you mentioned during our standup meeting yesterday morning. As we discussed, I have gone ahead and enabled two-step verification on your BlueMedia employee account so you can finally log into the new analytics dashboard without being blocked by the firewall rules we rolled out last week.\n\nWhen you attempt to sign in today, the system will prompt you for a secondary authentication code. Your temporary six-digit verification code is 482910. Please enter this code exactly as shown when the login screen asks for your two-step token. Keep in mind that this specific code will expire at midnight tonight, 11 March 2020, so you must complete your initial login setup before then. If you miss this window, you will need to reach out to me directly to generate a fresh one.\n\nOnce you are successfully logged in, I highly recommend navigating to your profile settings and linking your mobile phone number for future verification prompts. This will save us both a lot of back-and-forth emails going forward, especially since the engineering team is planning to enforce mandatory two-factor authentication across all BlueMedia departments starting next Monday, 19 March 2020. After that deadline, anyone who has not configured their secondary login method will be temporarily locked out of the internal network until their setup is verified by an administrator.\n\nYou can find the full step-by-step guide for configuring your mobile authenticator on our internal documentation site here: https://www.bluemedia.de/docs/security/two-step-setup-guide. That page walks you through downloading the approved authenticator app and scanning your unique QR code. Please make sure you are connected to the corporate Wi-Fi when you access that link, as it is not available on external networks.\n\nIf you run into any trouble entering the code, or if the portal throws an unexpected error during the setup wizard, just reply to this email or ping me on Slack. I will be at my desk for the rest of the afternoon and can walk you through the process via screen share if needed.\n\nThanks for getting this sorted out quickly.\n\nBest regards,\nBrandon Ali\nSystems Administrator\nBlueMedia GmbH"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00409-5", "copies": 1, "upstream_ids": ["gen-00409-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:sms_phishing:none:mendeley-823:message_type", "suite": "spam", "family": "sms_phishing-e393fb7a489118ea7705", "state": {"channel": "sms", "message": "Wish i were with you now!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-823", "copies": 2, "upstream_ids": ["mendeley-823", "uci-739"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2:is_spam", "suite": "spam", "family": "spamassassin-9303ed5da1fe4e373ad3", "state": {"channel": "email", "message": "Sorting\n\nBut there is a problem with making changes outside of exmh: the .xmhcache files\ndon't get updated. This is dangerous; I have to remember to re-scan each folder\nI enter. Is there a safeguard for this?\n\n--\nrick"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2", "zefang-16493", "ealvaradob-15555"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00112-2:is_spam", "suite": "spam", "family": "generated_legit_email-8678821c37f7734fc614", "state": {"channel": "email", "message": "Maplewood Community Garden – Seed Packet Order Dispatched\n\nMAPLEWOOD COMMUNITY GARDEN\n\nHi Robert Bailey,\n\nYour summer seed packet order has been dispatched today.\n\nTracking Number: MCG-2016-88472\nCarrier: Quartz Logistics\nPhone: (614) 555-0193\n\n[ Track My Shipment ]\n\nThank you for supporting our community garden!\n\nMaplewood Community Garden\n412 Elm Street, Maplewood, NJ 07040\nPrivacy Policy: see attached or call us."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00112-2", "copies": 1, "upstream_ids": ["gen-00112-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:zefang_phishing_email:none:zefang-6637:is_spam", "suite": "spam", "family": "zefang_phishing_email-7cc6793d75ec808a2a18", "state": {"channel": "email", "message": "ideas vince : i may have neglected to send you the attached last week . enron is one of a handful of institutions which could facilitate an entire transaction of the kind described here . this is simply your old production payment concept applied to corporate finance . best , - pressrelease . pdf - monetization . pdf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6637", "copies": 2, "upstream_ids": ["zefang-6637", "ealvaradob-6361"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-18256:is_spam", "suite": "spam", "family": "zefang_phishing_email-7d0a90c4536f3adde4c1", "state": {"channel": "email", "message": "meeting deadlines dear eugenio , i did not want to say this when everyone else was around , but i am concerned that no attempt was made to meet the deadline of this morning that we discussed last friday . ( to decide on a name for the database ) . only maria teresa had her information to me this am as requested . the deadline could have been easily met by working diligently this morning , but jennifer did not come in until 8 : 30 and mt until 8 : 15 . i thought we had discussed the urgency of this - to have something to present at the 10 am meeting . we need to discuss this to ensure it does not happen again . best regards shona"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18256", "copies": 2, "upstream_ids": ["zefang-18256", "ealvaradob-17170"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d:is_spam", "suite": "spam", "family": "spamassassin-393b14eafdcac3492703", "state": {"channel": "email", "message": "Founding Fathers on Religion\n\nSome interesting quotes...\n\nhttp://www.postfun.com/pfp/worbois.html\n\nThomas Jefferson:\n\n\"I have examined all the known superstitions of the word, and I do not\nfind in our particular superstition of Christianity one redeeming feature.\nThey are all alike founded on fables and mythology. Millions of innocent\nmen, women and children, since the introduction of Christianity, have been\nburnt, tortured, fined and imprisoned. What has been the effect of this\ncoercion? To make one half the world fools and the other half hypocrites;\nto support roguery and error all over the earth.\"\n\nSIX HISTORIC AMERICANS,\nby John E. Remsburg, letter to William Short\nJefferson again:\n\n\"Christianity...(has become) the most perverted system that ever shone on\nman. ...Rogueries, absurdities and untruths were perpetrated upon the\nteachings of Jesus by a large band of dupes and importers led by Paul, the\nfirst great corrupter of the teaching of Jesus.\""}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00701.4c1a296394c3afdcb38020e8283d1a5d", "zefang-101", "ealvaradob-99"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2801:message_type", "suite": "spam", "family": "sms_phishing-6135d6422aec7db149fd", "state": {"channel": "sms", "message": "Download as many ringtones as u like no restrictions, 1000s 2 choose. U can even send 2 yr buddys. Txt Sir to 80082 £3"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2801", "copies": 2, "upstream_ids": ["mendeley-2801", "uci-2352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3018:is_spam", "suite": "spam", "family": "sms_phishing-c337b135c28a6f9e5b7d", "state": {"channel": "sms", "message": "Detroit. The home of snow. Enjoy it."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3018", "copies": 2, "upstream_ids": ["mendeley-3018", "uci-4242"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4520:message_type", "suite": "spam", "family": "sms_phishing-b4717e64d2c76545d7b4", "state": {"channel": "sms", "message": "Sorry, I will call later..I am in a meeting.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4520", "copies": 1, "upstream_ids": ["mendeley-4520"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2023#184:is_spam", "suite": "spam", "family": "nazario-1d873e896565198b49c6", "state": {"channel": "email", "message": "Some information seems to be inaccurate #203997\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 13809892This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#184", "copies": 1, "upstream_ids": ["phishing-2023#184"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00209-2:message_type", "suite": "spam", "family": "generated_legit_email-c7dafae2dd3ff11743f8", "state": {"channel": "email", "message": "Your Charity Gala Tickets Are Confirmed\n\nHi Betty,\n\nI am writing to confirm that your ticket purchase for the Union Design Annual Charity Gala on January 22, 2015, has been processed successfully. We charged £150.00 to your Barclays account ending in 8832.\n\nYou have reserved two general admission tickets under the name Betty Kelly. The event will take place at the Grand Hall in London starting at 7:00 PM. Dinner and drinks are included in the price. Please bring a printed copy of this email or show it on your phone at the door so our team can scan you in quickly.\n\nIf you need to transfer these tickets to a colleague or update your dietary requirements before the event, you can manage your booking by visiting https://www.uniondesign.de/events/gala2015. I highly recommend doing this soon, as the guest list closes on January 10, 2015.\n\nThank you for supporting our charity initiative this year. We raised over £20,000 last time, and your contribution helps us continue funding local youth programs across the region. We really appreciate it.\n\nBest regards,\nRyan Green\nEvents Coordinator\nUnion Design\ndmitchell@btinternet.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00209-2", "copies": 1, "upstream_ids": ["gen-00209-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-3395:is_spam", "suite": "spam", "family": "sms_phishing-bd44bdbec3662db90d2d", "state": {"channel": "sms", "message": "Aight what time you want me to come up?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3395", "copies": 2, "upstream_ids": ["mendeley-3395", "uci-3042"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2769:message_type", "suite": "spam", "family": "sms_phishing-01281a67be428e6263f0", "state": {"channel": "sms", "message": "U r too much close to my heart. If u go away i will be shattered. Plz stay with me."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2769", "copies": 2, "upstream_ids": ["mendeley-2769", "uci-1312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#219:is_spam", "suite": "spam", "family": "nazario-4d6a7643adb32a96c3d1", "state": {"channel": "email", "message": "Subject: Attention anna\n\nWe're writing to let you know that we detected somethingunu force sual about recent sign-in to yourCh continue aseOn north linedep weight ositac can count. To help keep yourac deal count safe, werequ bring ire an extra securitychalle human nge we can help you take corrective action. Please log on to yourCh manage aseOn process lineac appear count by clicking on Dear Customer: Recently, there's beenact sort ivity in yourac eight count that seemsunu professional sual Compared to your Normalac audience count activities..What do you need to do? Please visit the following link tocon task firm yourac pull count information. Con performance firm Identity Note: This may have happened because you're using a device you don't usually use or you cleared the data on your browser. Thank you for helping us keep yourac available count safe. Sincerely, On help lineBa newspaper nking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#219", "copies": 1, "upstream_ids": ["phishing-2023#219"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-4769:is_spam", "suite": "spam", "family": "zefang_phishing_email-5fe38aff309cfd7c1966", "state": {"channel": "email", "message": "meter 980070 ; oct 99 daren , there are two purchases at the above referenced meter for the 28 th of oct . , 2010 . gas flowed without a purchase from the 29 th through the 31 st . please let me know if we should be paying the swing parties ( deal ticketsl 17738 and 120359 ) , or how i should allocate the volume . thank you for your time and patience ! ! mary jane ext . 35251"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4769", "copies": 2, "upstream_ids": ["zefang-4769", "ealvaradob-4562"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5095:is_spam", "suite": "spam", "family": "sms_phishing-ad1612386dc96bbeb4e9", "state": {"channel": "sms", "message": "I'm home, my love ... If your still awake ... *loving kiss*"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5095", "copies": 2, "upstream_ids": ["mendeley-5095", "uci-4723"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3919:is_spam", "suite": "spam", "family": "sms_phishing-1998f66696a3aa04d861", "state": {"channel": "sms", "message": "But my family not responding for anything. Now am in room not went to home for diwali but no one called me and why not coming. It makes me feel like died."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3919", "copies": 2, "upstream_ids": ["mendeley-3919", "uci-2744"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#154:message_type", "suite": "spam", "family": "nazario-73c376465efedd4a025f", "state": {"channel": "email", "message": "Place or Update Credit Card on File\n\neBay treats your personal information with utmost care.\nThis eBay notice was sent to you based on your eBay account preferences and in accordance with our Privacy Policy.\n\nPlace or Update Credit Card On File\n\nDear eBay user,\n\nDuring our regularly schedule account maintenance and verification we have detected a slight error in your billing information on file with eBay. This might be due to either following reasons:\n\n- A recent change in your personal information (i.e. change of address)\n- Submitting invalid information during the initial sign up process.\n- An inability to accurately verify your selected option of payment due an internal error within\nour processors.\n\nPlease sign in to your eBay account and update your billing information:\n\nhttp://signin.ebay.com/eBayISAPI.dll?SignIn&ssPageName=h:h:sin:US\n\nIf your account information is not update, your ability to sell or bid on eBay will become restricted.\n\nThank you,\neBay Billing Department\n\nVisit our Privacy Policy and User Agreement if you have any questions.\n\nProtect yourself from Spoof (fake) e-mails. Learn More\n\nCopyright © 2014-2024 eBay Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are trademarks of eBay Inc."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#154", "copies": 1, "upstream_ids": ["phishing2.mbox#154"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00548-5:is_spam", "suite": "spam", "family": "generated_legit_email-ca201f8e716f13449e8d", "state": {"channel": "email", "message": "Question about your listed item: Vintage Desk Lamp\n\nHi Scott Flores,\n\nI am interested in buying the vintage desk lamp you posted on the Trident Systems Community Marketplace. A few quick questions:\n\n1. Is the original power cord still intact and safe to use?\n2. Would you accept $35.00 instead of the listed $45.00?\n3. Could I pick it up from your office before 18 April 2013?\n\nPlease let me know.\n\nThanks,\nRachel Adams"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00548-5", "copies": 1, "upstream_ids": ["gen-00548-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:sms_phishing:none:mendeley-2897:is_spam", "suite": "spam", "family": "sms_phishing-f799641df85f50299e12", "state": {"channel": "sms", "message": "Nice.nice.how is it working? Tell me once u r done"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2897", "copies": 1, "upstream_ids": ["mendeley-2897"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2539:message_type", "suite": "spam", "family": "sms_phishing-6baf99e48dcf16b45b03", "state": {"channel": "sms", "message": "FREE>Ringtone! Reply REAL or POLY eg REAL1 1. PushButton 2. DontCha 3. BabyGoodbye 4. GoldDigger 5. WeBeBurnin 1st tone FREE and 6 more when u join for £3/wk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2539", "copies": 2, "upstream_ids": ["mendeley-2539", "uci-3921"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2:message_type", "suite": "spam", "family": "spamassassin-9303ed5da1fe4e373ad3", "state": {"channel": "email", "message": "Sorting\n\nBut there is a problem with making changes outside of exmh: the .xmhcache files\ndon't get updated. This is dangerous; I have to remember to re-scan each folder\nI enter. Is there a safeguard for this?\n\n--\nrick"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00983.42eb62cc4057d03dfd3eb1c8afc270e2", "zefang-16493", "ealvaradob-15555"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_spam:none:uci-3392:is_spam", "suite": "spam", "family": "sms_spam-86344b4c7e98519581f0", "state": {"channel": "sms", "message": "I'm in a meeting, call me later at"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-3392", "copies": 3, "upstream_ids": ["uci-3392", "uci-4633", "uci-590"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5890:message_type", "suite": "spam", "family": "sms_phishing-cc64af943d07b8293689", "state": {"channel": "sms", "message": "Double Mins & Double Txt & 1/2 price Call MobileUpd8 for the very latest offers. 08000839402 or call2optout/LF56"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5890", "copies": 1, "upstream_ids": ["mendeley-5890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5933:message_type", "suite": "spam", "family": "sms_phishing-91a292e29d14ebfc7373", "state": {"channel": "sms", "message": "We spend our days waiting for the ideal path to appear in front of us.. But what we forget is.. \"paths are made by walking.. not by waiting..\" Goodnight!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5933", "copies": 2, "upstream_ids": ["mendeley-5933", "uci-1382"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2:message_type", "suite": "spam", "family": "spamassassin-d223622360a78d9332ea", "state": {"channel": "email", "message": "What you need for iPaq wireless connectivity (Tech Update)\n\nTech Update Today\n\nVITAL SIGNS FOR JULY 12, 2014\n\nDavid Berlind\n\nWhat you need for iPaq wireless connectivity\n\nA public health organization has several iPaqs on order, but\nnow they want to add WAN connectivity to those PDAs. What\nexactly do they need to do? I'm so glad they asked.\n\nHere's everything you need to know\n\nLatest from ZDNet News\n\nSecurity holds back Web services\n\nChina: Bursting with brainpower\n\nMicron cranks handle on next-gen memory\n\nJapan chipmakers team on 'SoCs' chip\n\nHP tech exec jumps ship to EMC\n\nHoles put Linux handheld at risk\n\nMicrosoft bids to be creme de la CRM\n\nMore Enterprise News\n\nFarber's Picks\n\nDan\nFarber\n\nDesperately seeking chief ethics officer\n\nGreedy CEOs are making a bad name for the profession--and corporate America. The title chief executive officer now\nconjures up an image of someone with questionable ethics, a\ngreedy disposition, and a disregard for shareholders. If\ngovernment-proposed remedies can\"t adequately address this\nethics crisis, disgruntled shareholders surely will.\n\nRead my full commentary\n\nOracle challenges Microsoft anew\n\nOracle is taking on Microsoft's Exchange with a Collaboration Suite that includes e-mail, calendar and messaging software. Oracle has pitched its products as Microsoft replacements before -- with spectacularly little success. (At Comdex last November, the database giant offered its 9i and app-server software as a more secure alternative to Exchange.) What's different this time? Oracle is selling a separate collaboration product that, says CEO Larry Ellison, is cheaper than rival offerings. Plus: Companies facing a deadline to join Redmond's new licensing program may be more open to non-Micrososft alternatives.\n\nRead the full story\n\nAlso: Ellison shrugs a No. 2 exec\n\nStandards stalled over royalty disputes\n\nThe World Wide Web Consortium is bracing for a vote next week that could decide once and for all how the standards body will handle patented technology that comes with royalties attached. If patented technology is rejected, the vote could mark a major shift in the Internet standards battle. Major intellectual property holders such as Microsoft and IBM have a lot at stake.\n\nRead the full story\n\nMicrosoft's OpenGL claim shakes graphics world\n\nSurvey: Execs rethinking IT budgets\n\nGrowing numbers of Fortune 1000 CIOs are re-evaluating their IT budgets and spending plans for coming months, according to a recent survey. Morgan Stanley finds that CIOs' top priorities include security software, application integration and Windows 2012/XP upgrades for the desktop. Integration challenges were slowing the rate of application adoption for 38 percent of those surveyed.\n\nRead more survey results\n\nOh look, a new bug in Outlook and IE\n\nA Danish security expert has warned of a security flaw in Microsoft's ActiveX scripting component that allows network attackers to peek at cookies, read files and even run programs on a victim's PC. The expert recommended that users disable ActiveX in IE and Outlook, at least until Microsoft releases a patch.\n\nRead the full story\n\nAlso: Outlook plug-in opens PCs to hackers\n\nWrite me at sofia_brown@orange.fr\n\nBack to top\n\nAlso on Tech Update Today\n\nCOMMENTARY\n\nMixing Unix and Windows\n\nMost enterprises have a mix of Windows and Unix systems. Larry Seltzer explores the options for interoperating those systems.\n\nQ&A\n\nMarcelo Tosatti: The future is Linux\n\nThe maintainer of the Linux kernel talks about new opportunities, defends UnitedLinux and tells why copyright laws keep him from entering the U.S.\n\nPREVIOUSLY ON TECH UPDATE TODAY\n\nROI calculators don't tell the whole TCO story\n\nIn response to David Berlind's \"Dell's formula for ROI,\" TCO\nconsultant Tom Rankin warns that \"99 percent of TCO tools do\nnot look at the 'total' but may look at whatever parameters\nsuit the vendor to sell their products/services.\"\n\nCrucial Clicks--products worth looking at\n\nNetworking\n\nFind the right home network\nChoosing the right technology to match your networking needs is vital. Here's help in making the choice.\n\nRead reviews\n\nMost Popular Products\n\nNetworking\n\n1. Linksys EtherFast wireless AP\n\n2. Linksys EtherFast router\n\n3. Siemens SpeedStream router\n\n4. Wireless 802.11B Router\n\n5. NetGear HE102 802.11a wireless AP\n\nMore popular networking products\n\nElsewhere on ZDNet\n\nVeritas CEO Gary Bloom speaks out at Tech Update.\nCheck out these personal laser printer picks at ZDNet Reviews.\nSee how WAP works and how you should use it at Builder.com.\nNeed a new job? Find one today in ZDNet's Career Center.\nCIOs talk out about the future of IT at CNET News.com.\n\nSign up for more free newsletters from ZDNet\n\nThe e-mail address for your subscription is tbrown@orange.fr\n\nManage\nMy Subscriptions | FAQ |\nAdvertise\n\nHome |eBusiness | Security | Networking | Applications | Platforms | Hardware | Contact us\n\nCopyright 2014 CNET Networks, Inc. All rights reserved. ZDNet is a registered service mark of CNET Networks, Inc."}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00056.c17c95457d7524edd8a72b6de455c4d2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2022#224:is_spam", "suite": "spam", "family": "nazario-ac1f7d93b05c6e4eefbb", "state": {"channel": "email", "message": "Action Required: UPS Exception Notification\n\nHello,\n\nWe inform you that your package 1Z6R18V12048992414 is in transit on our logistics platform and is awaiting instruction from you.\n\nPlease confirm the payment of the delivery fee ($2,49) and the delivery address of the package by clicking here.\n\nFollow-up message from UPS:\n\nOrder: o-yf3z8xy looking forward to joining you|\n\nShipping information\n\nTracking number :\n\n1Z6R18V12048992414\n\nUPS Service :\n\nUPS STANDARD\n\nPackage weight :\n\n3.90 KGS\n\nOrder number:\n\nN87654690\n\nDownload the UPS Mobile Application\n\n©2016 United Parcel Service of America, Inc. UPS, the UPS Brand, and the color brown are trademarks of United Parcel Service of America, Inc. All rights reserved.\n\nAll trademarks, trade names or service marks appearing in connection with UPS services are the property of their respective owners.\n\nPlease do not reply directly to this message. UPS will not receive your message.\n\nView the UPS Privacy Policy\n\nIf you have any questions, please visit our help and support center"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#224", "copies": 1, "upstream_ids": ["phishing-2022#224"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5308:is_spam", "suite": "spam", "family": "sms_phishing-2315858696bed84ccae4", "state": {"channel": "sms", "message": "U still going to the mall?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5308", "copies": 2, "upstream_ids": ["mendeley-5308", "uci-177"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_spam:none:uci-4200:is_spam", "suite": "spam", "family": "sms_spam-e5276c26bb48d59a94d1", "state": {"channel": "sms", "message": "cmon babe, make me horny, *turn* me on! Txt me your fantasy now babe -) Im hot, sticky and need you now. All replies cost £1.50. 2 cancel send STOP"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-4200", "copies": 1, "upstream_ids": ["uci-4200"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:spam"], "datasets": ["sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848:message_type", "suite": "spam", "family": "spamassassin-eedd5835f2210bec32ed", "state": {"channel": "email", "message": "ok, in a galeon browser window i open a new window by middle-clicking\na link. the window comes up under my mouse. however, keys i type go to\nthe first window. anyone else have this problem, and did they solve it?\ni use sawfish as my wm with sloppy focus.\n\nkevin\n\n--\ndeborah40@comcast.net that a believer is happier than a skeptic is no more to\nfork()'ed on 37058400 the point than the fact that a drunken man is happier\nmeatspace place: home than a sober one. the happiness of credulity is a\nhttp://ie.suberic.net/~kevin cheap & dangerous quality -- g.b. shaw\n\n--"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00297.cd323132d57e5cb2e8599f69f6ff2848", "zefang-3464", "ealvaradob-3323"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12527:is_spam", "suite": "spam", "family": "zefang_phishing_email-9758d0a209c71dd7e060", "state": {"channel": "email", "message": "On Thu, Aug 08, 2024 at 08:00:00PM -0700,\nJehan is thought to have said:> Actually you would be inaccurate in your action in a way. The correct\nwith you. It makes no difference what the list is, whether it's run by ZDNet\nor Empire Towers, whether it uses open proxies and relays or a MLM like\nmailman or Lyris. In either case I was added to a list against my will to\nthe list' any more than I should have to send email with the subject REMOVE\nto kwame.gray@horizonworks.com or go to http://www.centralremovelist.com/\ncommitting an inherently untrustworthy act. Yet someone's TES would likely\nfall because of it.> Razor is designed for the betterment of most people, i.e. the majority,\nTabor J. Wells jcooper@mail.com\nFsck It! Just another victim of the ambient morality\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12527", "copies": 2, "upstream_ids": ["zefang-12527", "ealvaradob-11864"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3791:is_spam", "suite": "spam", "family": "sms_phishing-b820aaa220d1ed3e784c", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 13 jan AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play. 8.Neighbour: We went 2 a marriage. who is the murderer?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3791", "copies": 1, "upstream_ids": ["mendeley-3791"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3306:is_spam", "suite": "spam", "family": "sms_phishing-8d384fa0bd0c36428ffb", "state": {"channel": "sms", "message": "Ugh y can't u just apologize, admit u were wrong and ask me to take u back?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3306", "copies": 2, "upstream_ids": ["mendeley-3306", "uci-2031"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00387-5:is_spam", "suite": "spam", "family": "generated_legit_email-1c26976c394e31bb5501", "state": {"channel": "email", "message": "Question about your Quartz Consulting project listing\n\nQUARTZ CONSULTING NETWORK\n\nHi Fatima Baker,\n\nA potential client has contacted you about your consulting services.\n\nClient: Kenji Brown\nProject: IT Infrastructure Audit\nProposed Budget: $4,500.00\n\nKenji Brown writes:\n\"Hello Fatima, I found your profile on the Quartz directory. We need someone to audit our server setup by late February. Are you available to discuss this week? Please let me know your hourly rate.\"\n\nReply to Kenji Brown\n\nView Message: https://www.quartzconsulting.net/messages/49201\n\nThanks,\nThe Quartz Consulting Team\n\nQuartz Consulting Inc.\n1440 Broadway, New York, NY 10018\nPrivacy Policy: https://www.quartzconsulting.net/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00387-5", "copies": 1, "upstream_ids": ["gen-00387-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "marketplace message from a buyer or seller about an item"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502:message_type", "suite": "spam", "family": "spamassassin-6abe80dd0e72bbc8a0ca", "state": {"channel": "email", "message": "Thank you for your response I will verify that I have taken these steps\nthat you have metioned but as well I would like to stop mail from\nentering my mail boxes and need the info about how to integrate with\nqmail in newbie terms if that is possible. Thank you again for your\nhelp\n\nMichael Humphries\nAlphacomm.net\nNetwork Administrator\n906-639-3500\n877-450-3500\nivan.bailey@comcast.net\nwww.alphacomm.net\n\nHiggins\n\ni don't think you get exactly what razor does. razor is used to stop\nspam\nfrom entering a mail box, not to leave a mail server. altough you could\nprobably use it that way i doubt it would do much good (if the mail is\nleaving the server, it most likely is a new spam and wouldn`t get caught\nby\nrazor)\n\nyou might want to look into dealing with those troublesome users\ndirectly.\nand (as you`ve probably been screamed at a lot of times) close your open\nrelay if you have one, there are plenty of sites that tells you how\n(most\nRBL sites)\n--\nDaniel Higgins\nNetcommunications Inc.\nAdministrateur Système / Network Administrator\nTel: 450-346-3401\nFax: 450-346-3587\ndavid96@horizoncapital.org\nhttp://www.netc.net\n\nurgent\nstop\n\nfor high-tech professionals. Search and apply for tech jobs today!\nhttp://seeker.dice.com/seeker.epl?rel_code1\n\nfor high-tech professionals. Search and apply for tech jobs today!\nhttp://seeker.dice.com/seeker.epl?rel_code1"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00631.3f9351806af9a88f93169ee3416e4502"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2664:message_type", "suite": "spam", "family": "sms_phishing-6c4f0a6b26a8a43f8749", "state": {"channel": "sms", "message": "Hi frnd, which is best way to avoid missunderstding wit our beloved one's?"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2664", "copies": 2, "upstream_ids": ["mendeley-2664", "uci-168"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-69:is_spam", "suite": "spam", "family": "sms_phishing-d4df68cfb848ffe87070", "state": {"channel": "sms", "message": "You need to get up. Now."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-69", "copies": 2, "upstream_ids": ["mendeley-69", "uci-3936"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00646-0:is_spam", "suite": "spam", "family": "generated_legit_email-dcac9b4bdff92927d567", "state": {"channel": "email", "message": "Interview Invitation – Content Strategist Position\n\nDear Pamela Nelson,\n\nThank you for your recent application to North Media Group for the Content Strategist position. We have reviewed your materials and were impressed by your background in digital publishing. We would like to invite you to an interview to discuss the role further.\n\nPlease review the following details regarding your upcoming interview:\n\n1. Interview Date: Wednesday, 12 June 2008\n2. Interview Time: 10:30 AM (BST)\n3. Location: North Media Group Headquarters, 45 Fleet Street, London EC4Y 1DP\n4. Interviewer: Noah Byrne, Head of Editorial Operations\n\nTo ensure the process runs smoothly, please complete the numbered steps below before your arrival:\n\n1. Confirm your attendance by replying directly to this email no later than Friday, 7 June 2008.\n2. Prepare a brief portfolio presentation (maximum ten minutes) highlighting three campaigns you have managed.\n3. Bring two forms of photographic identification for our building security check-in.\n4. Arrive fifteen minutes early so we can offer you tea or coffee before the meeting begins.\n\nDuring the interview, we will cover the following topics:\n- Your experience with SEO-driven content planning\n- Team leadership and cross-department collaboration\n- Salary expectations and your earliest available start date\n\nWe are very much looking forward to meeting you in person and learning more about how your skills align with our current projects at North Media Group. If you require any reasonable adjustments for the interview or need directions to our office, please do not hesitate to reach out.\n\nYou can contact me directly by phone at +44 20 7946 0123 during standard business hours (Monday to Friday, 9:00 AM to 5:00 PM).\n\nWarm regards,\nNoah Byrne\nHead of Editorial Operations\nNorth Media Group\n+44 20 7946 0123"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00646-0", "copies": 1, "upstream_ids": ["gen-00646-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-4265:message_type", "suite": "spam", "family": "sms_phishing-37b58ec81161585d9c73", "state": {"channel": "sms", "message": "There is a first time for everything :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4265", "copies": 2, "upstream_ids": ["mendeley-4265", "uci-5411"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4043:is_spam", "suite": "spam", "family": "sms_phishing-0e1f7b4ad088ba67900f", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-7908807538 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4043", "copies": 1, "upstream_ids": ["mendeley-4043"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5393:message_type", "suite": "spam", "family": "sms_phishing-5c5146c8d15fa8055577", "state": {"channel": "sms", "message": "(6171-M) RMO.00 AEON MY Rs. 4399 was charged on your card num**97 at LAZO-KLIA2 on 19/10/2013. Call 0327884425 if you did NOT perform this TXN code356931"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5393", "copies": 1, "upstream_ids": ["mendeley-5393"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5375:message_type", "suite": "spam", "family": "sms_phishing-e06453e2191089de14a8", "state": {"channel": "sms", "message": "I can't have a friend in the house because i'm embarassed that everything is scattered on the floor. I think we should work on it."}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5375", "copies": 1, "upstream_ids": ["mendeley-5375"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4211:is_spam", "suite": "spam", "family": "sms_phishing-002b52b5f68fa502e846", "state": {"channel": "sms", "message": "Uh, heads up we don't have THAT much left"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4211", "copies": 2, "upstream_ids": ["mendeley-4211", "uci-1896"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-966:is_spam", "suite": "spam", "family": "sms_phishing-5193b8f2e13e008738d9", "state": {"channel": "sms", "message": "I know you are. Can you pls open the back?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-966", "copies": 2, "upstream_ids": ["mendeley-966", "uci-144"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1672:is_spam", "suite": "spam", "family": "sms_phishing-fb7556b41ef040185fa0", "state": {"channel": "sms", "message": "Ok going to sleep. Hope i can meet her."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1672", "copies": 2, "upstream_ids": ["mendeley-1672", "uci-3135"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-793:is_spam", "suite": "spam", "family": "sms_phishing-f9d58353ceb8756093d1", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\" Love u dear"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-793", "copies": 1, "upstream_ids": ["mendeley-793"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-718:is_spam", "suite": "spam", "family": "sms_phishing-85599350121c03f04bc6", "state": {"channel": "sms", "message": "I've reached already."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-718", "copies": 2, "upstream_ids": ["mendeley-718", "uci-4730"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4224:is_spam", "suite": "spam", "family": "zefang_phishing_email-989806c83eff4783592e", "state": {"channel": "email", "message": "northeast spring membership mixer february 4 , 2022 nesa members : attached is our northeast spring membership mixer to be held at macmenamin ' s south street seaport - nyc ! > > other great upcoming events : feb . 20 , 2022 - storage ecomonics brown bag mar . 12 - 13 , 2022 - nominations thru allocation technical training mar . 18 , 2022 - charity golf tournament please visit our website www . nesanet . org for more information or you can call us direct ( 713 ) 856 - 6525 . thanks ! lana moore director or education nesa"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4224", "copies": 2, "upstream_ids": ["zefang-4224", "ealvaradob-4045"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#204:is_spam", "suite": "spam", "family": "nazario-a27c56e2010ecb95b8ec", "state": {"channel": "email", "message": "Automated message do not Reply 19/07/2021\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 19355562This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#204", "copies": 1, "upstream_ids": ["phishing-2023#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00008-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-c8161ea42caaab0a51dd", "state": {"channel": "email", "message": "Software Licence Renewal Notice\n\nSummit Works Ltd\nInvoice Number: SW-8834\nRecipient: Nicole Ivanov (jivanov@pioneerlabs.com)\nItem: Professional Office Suite Licences (5 seats)\nTotal: €119.00\nAction Required By: 18 August 2019\n\nDear Nicole,\n\nSecure discounted software licences before the deadline above. Visit https://www.summitworks.biz/licences to complete your order today.\n\nKind regards,\nKenji Anderson\nSales Director, Summit Works Ltd"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00008-5", "copies": 1, "upstream_ids": ["genu-00008-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "advertising for cheap software licences", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-61:is_spam", "suite": "spam", "family": "sms_phishing-ee801d9a8a387dc28220", "state": {"channel": "sms", "message": "Sorry, have urgent work.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-61", "copies": 1, "upstream_ids": ["mendeley-61"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_unwanted_email:none:genu-00104-2:message_type", "suite": "spam", "family": "generated_unwanted_email-42616a914d3f1ffe9dd9", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected on your account\n\nDear Lars White,\n\nWe detected a sign-in to your Meridian Trust account from an unfamiliar device in Lagos, Nigeria, on 1 September 2012. Your account has been temporarily locked for your protection.\n\nPlease verify your identity immediately to restore access: https://www.quartznetworks32.info/verify-session\n\nKind regards,\nKwame Wilson\nFraud Prevention Team\nQuartz Financial Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00104-2", "copies": 1, "upstream_ids": ["genu-00104-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00459-2:message_type", "suite": "spam", "family": "generated_legit_email-9ee06b6a105de0ba9f13", "state": {"channel": "email", "message": "Action Required: Q1 Design Review Document\n\nReview Request\nRequester: Frank Park\nProject: Falcon Design Platform Update\nDocument: Q1_Review_Draft_v3.pdf\nDeadline: Friday, 6 February 2016\n\nHi Priya,\n\nPlease review the attached Q1 design document and add your comments before the deadline above. We need your approval to proceed with the next sprint.\n\nThanks,\nFrank Park\nFalcon Design"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00459-2", "copies": 1, "upstream_ids": ["gen-00459-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:sms_phishing:none:mendeley-4275:is_spam", "suite": "spam", "family": "sms_phishing-1d7b85bca8591f1f82c1", "state": {"channel": "sms", "message": "Dear good morning how you feeling dear"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4275", "copies": 2, "upstream_ids": ["mendeley-4275", "uci-3293"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-4536:is_spam", "suite": "spam", "family": "zefang_phishing_email-aa94972d8f4e19ced49f", "state": {"channel": "email", "message": "call for genetic programming book proposals call for book proposals for kluwer book series on genetic programming kluwer academic publishers announces the genetic programming book series genetic programming is a technique for automatically synthesizing computer programs to solve problems . the kluwer book series on genetic programming will cover applications of genetic programming , theoretical foundations of genetic programming , technique extensions , and implementation issues . it be the first collection of monographs , edited collections , and advanced texts to cover this rapidly growing field . in order to publish material that is timely and reflects the state of the art , the series will focus on books of relatively narrow scope and moderate length and will feature a rapid publication schedule . the first book of the series , langdon 's genetic programming and data structures : genetic programming + data structures = automatic programming ! has already been published . topics may include , but are not limited to design , control , classification , system identification , data mining , pattern recognition and image analysis , data and image compression , evolvable machine language , evolvable hardware , and automatic programming of multi-agent and distributed systems . prospective authors : if you have an idea for a book which would fit in this series , we would welcome the opportunity to review your proposal . should you wish to discuss any potential project further or receive specific information regarding our book proposal requirements , please contact either john koza or scott delman . pleas enclose a short biography with your proposal . john r . koza consulting editor section on medical informatics department of medicine school of medicine medical school office building stanford university 94305-5479 usa e-mail : koza @ genetic-programming . org phone : 650-941 - 0336 fax : 650-941 - 9430 scott delman senior publishing editor kluwer academic publishers 101 philip drive assinippi park norwell , ma 02146 phone : 781-871 - 6311 ext . 299 fax : 781-871 - 7507 email : sdelman @ wkap . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4536", "copies": 2, "upstream_ids": ["zefang-4536", "ealvaradob-4336"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1256:message_type", "suite": "spam", "family": "sms_phishing-78319c0379b8acc441c7", "state": {"channel": "sms", "message": "Y de asking like this."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1256", "copies": 2, "upstream_ids": ["mendeley-1256", "uci-4610"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1598:is_spam", "suite": "spam", "family": "sms_phishing-24116e37d49f05d044a9", "state": {"channel": "sms", "message": "Double your mins & txts on Orange or 1/2 price linerental - Motorola and SonyEricsson with B/Tooth FREE-Nokia FREE Call MobileUpd8 on 08000839402 or2optout/HV9D"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1598", "copies": 2, "upstream_ids": ["mendeley-1598", "uci-2343"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00652-3:message_type", "suite": "spam", "family": "generated_legit_email-a2de4058bfc87f987791", "state": {"channel": "email", "message": "Donation Receipt - Annual Charity Drive\n\nFalcon Media Internal Giving Campaign\n\nHi Patricia,\n\nThank you for your generous contribution to our annual charity drive. Your donation helps us continue supporting local community projects. Below are the details of your gift for your tax records.\n\nDonor Name: Patricia Williams\nReceipt Number: FD-2021-0892\nDate of Donation: 4 November 2021\nAmount: €150.00\nCharity Fund: Falcon Media Community Trust\nTax Deductible: Yes\n\nYou can download a printable PDF version of this receipt from our internal portal at https://www.falconmedia.de/receipts/0892.\n\nWe truly appreciate your support this year. If you need any further documentation for your tax filing, please do not hesitate to reach out to me directly.\n\nWarm regards,\nPriya Patel\nCommunity Outreach Manager\nFalcon Media\nPhone: +49 30 555 0192"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00652-3", "copies": 1, "upstream_ids": ["gen-00652-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:sms_phishing:none:mendeley-3362:is_spam", "suite": "spam", "family": "sms_phishing-ff124046adcda3871cca", "state": {"channel": "sms", "message": "You can donate £2.50 to UNICEF's Asian Tsunami disaster support fund by texting DONATE to 864233. £2.50 will be added to your next bill"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3362", "copies": 2, "upstream_ids": ["mendeley-3362", "uci-4968"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5121:is_spam", "suite": "spam", "family": "sms_phishing-f992119afa12bc89e330", "state": {"channel": "sms", "message": "Wow. I never realized that you were so embarassed by your accomodations. I thought you liked it, since i was doing the best i could and you always seemed so happy about \"the cave\". I'm sorry I didn't and don't have more to give. I'm sorry i offered. I'm sorry your room was so embarassing."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5121", "copies": 2, "upstream_ids": ["mendeley-5121", "uci-53"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00006-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-35de27fd68fee4c998db", "state": {"channel": "email", "message": "End of Summer Sale – Up to 60% Off Furniture and Electronics\n\nDear Susan Walker,\n\nI hope this message finds you well. My name is Sophie Rossi, and I am writing on behalf of Sterling Home & Tech, a small independent retailer based in Portland, Oregon.\n\nWe are currently running our End of Summer clearance event, offering up to 60% off selected furniture, clothing, and electronics. Our stock includes oak dining tables, linen summer dresses, wireless headphones, and smart televisions from leading brands.\n\nBecause we are clearing out our warehouse for autumn inventory, these prices are available only until 25 August 2023. After that date, all remaining items will return to their original retail price.\n\nIf you would like to receive our full product catalogue or place an order, please reply directly to this email with your request, and I will send the details right away. You can also reach us by calling our shop at (503) 555-0172 during business hours.\n\nThank you for your time, and I hope you find something you love.\n\nWarm regards,\nSophie Rossi\nSales Coordinator\nSterling Home & Tech"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00006-0", "copies": 1, "upstream_ids": ["genu-00006-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk discount offer on electronics, clothes or furniture from an unknown company", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-13152:is_spam", "suite": "spam", "family": "zefang_phishing_email-81edeeb8858a0a15641c", "state": {"channel": "email", "message": "home address i am working on a charity event , a night in monte carol that benefits the stehlin foundation . believe it or not , i am not looking for a donation or as i call it \" begging for bucks \" , i simply would to extend an invitation to you and spouse . the event will be on friday , september 21 st at the four seasons hotel and will include gambling , jazz music , french cuisine and a chance to bid on some great auction items ( like trips to washington or las vegas , golf packages or an original thomas kinkaid painting with certificate and much more ! ) . i realize that some of you have a conflict , the tw customer meeting and will not be able to attend but , wanted to send an invitation should you plans change and you are looking for an exciting event to attend . when you have a moment , please forward me your home address via email . thanks . the stehlin foundation for cancer research has earned an international reputation for it ' s pioneering research and treatment programs in the fight against cancer . dr . john s . stehlin , jr . established the foundation in 1969 with a specific and urgent mission : to find and develop the most effective treatments for patients suffering from cancer in the shortest period of time ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13152", "copies": 2, "upstream_ids": ["zefang-13152", "ealvaradob-12441"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2268:is_spam", "suite": "spam", "family": "sms_phishing-48d85ce3c8c5d6387a54", "state": {"channel": "sms", "message": "No no. I will check all rooms befor activities that's required"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2268", "copies": 1, "upstream_ids": ["mendeley-2268"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-416:message_type", "suite": "spam", "family": "sms_phishing-900c43c65f076cf807d8", "state": {"channel": "sms", "message": "Wen did you get so spiritual and deep. That's great"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-416", "copies": 2, "upstream_ids": ["mendeley-416", "uci-5553"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#771:message_type", "suite": "spam", "family": "nazario-bb5a68bde82801b9f1db", "state": {"channel": "email", "message": "Your payment has been sent to sales@configure.us.dell.com\n\nDear PayPal Member,\n\nThis email confirms that you have paid\nDELL INSPERION E1505\n\n(sales@configure.us.dell.com)\n$449.99 USD using\nPayPal.\n\nThis credit card transaction will appear on your\nbill as \"PAYPAL\nDELL INSPERION E1505 Notebook Computer\n\n\".\n\nPayPal Shopping Cart Contents\n\nItem Name:\n\nDell Inspiron 1300 Notebook Computer for Business 400 MHz - Celeron - 1 MB RAM - 40 GB Hard Drive\n15.4\" Display - Windows XP Home\n\nQuantity:\n\n1\n\nTotal:\n\n$499.99\n\nCart Subtotal:\n\n$484.99 USD\n\nSales Tax:\n\n$15.00 USD\n\nCart Total:\n\n$499.99\n\nShipping Information\n\nShipping Info:\n\nBrandi Lee Skiles\n\n2700 Cross Point circle\n\nMatthews, North Carolina 28105\n\nUnited States\n\nAddress Status:\n\nUnconfirmed\n\nIf you haven't authorized this charge, click the link below to cancel the payment and get a full refund.\n\nDispute Transaction\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this e-mail. Mail sent to this address cannot be\nanswered. For assistance, log in to your PayPal account and choose the \"Help\" link in the footer of\nany page.\n\nPayPal Email ID PP133320"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#771", "copies": 1, "upstream_ids": ["phishing3.mbox#771"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00043-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-03f74abe48b2acc2f8ec", "state": {"channel": "email", "message": "Congratulations – You have been selected for the Atlas Labs Annual Prize Draw\n\nDear Cynthia Sanchez,\n\nI am delighted to inform you that your email address was selected as a winner in the Atlas Labs Annual Prize Draw held on 10 January 2016. You have been awarded a prize of £8,500.\n\nTo process your winnings and arrange the transfer to your bank account, we kindly ask you to provide your full name, residential address, and a small processing fee of £45. Please complete your claim by visiting https://www.atlaslabs58.co/claim-prize no later than 28 January 2016.\n\nShould you require any assistance during this process, please do not hesitate to call our claims department directly at +44 20 7946 0321.\n\nWe look forward to hearing from you soon and congratulate you once again on your win.\n\nWarmest regards,\nDonna Bennett\nAtlas Labs Promotions Team"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00043-1", "copies": 1, "upstream_ids": ["genu-00043-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a prize or lottery win notice that asks for personal details or a processing fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00611-4:message_type", "suite": "spam", "family": "generated_legit_email-4a4245bf1c7535a421d9", "state": {"channel": "email", "message": "Notice: Scheduled Water Maintenance - Unit 4B\n\nNORTH DESIGN PROPERTY MANAGEMENT\n================Building Maintenance Notice\n\nDear Jessica King,\n\nWater supply to Unit 4B will be shut off on Monday, 5 November 2020 from 9 AM to 1 PM for pipe repairs.\n\nView maintenance schedule\nhttps://www.northdesign.net/notices/water-maintenance-nov2020\n\nThank you,\nJennifer Lopez\nNorth Design Property Management\n\n(c) 2020 North Design Ltd. 44 Birch Street, Manchester M1 4BT.\nPrivacy Policy: https://www.northdesign.net/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00611-4", "copies": 1, "upstream_ids": ["gen-00611-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:sms_phishing:none:mendeley-174:is_spam", "suite": "spam", "family": "sms_phishing-87dc09121801bc216ac1", "state": {"channel": "sms", "message": "At 4. Let's go to bill millers"}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-174", "copies": 2, "upstream_ids": ["mendeley-174", "uci-1589"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3082:message_type", "suite": "spam", "family": "sms_phishing-c103ee04d140ab25e33b", "state": {"channel": "sms", "message": "We tried to call you re your reply to our sms for a video mobile 750 mins UNLIMITED TEXT free camcorder Reply or call now 08000930705 Del Thurs"}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": [0.0, 0.5, 0.5], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3082", "copies": 3, "upstream_ids": ["mendeley-3082", "mendeley-4788", "uci-3999"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4043:message_type", "suite": "spam", "family": "sms_phishing-0e1f7b4ad088ba67900f", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-7908807538 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4043", "copies": 1, "upstream_ids": ["mendeley-4043"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4393:message_type", "suite": "spam", "family": "sms_phishing-fd71d15dbd8acd3c8943", "state": {"channel": "sms", "message": "Sary just need Tim in the bollox &it hurt him a lot so he tol me!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4393", "copies": 2, "upstream_ids": ["mendeley-4393", "uci-2973"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1780:is_spam", "suite": "spam", "family": "sms_phishing-a5ac5c5038c2d8777335", "state": {"channel": "sms", "message": "Hey ! Don't forget ... You are MINE ... For ME ... My possession ... MY property ... MMM ... *childish smile* ..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1780", "copies": 2, "upstream_ids": ["mendeley-1780", "uci-5289"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#875:is_spam", "suite": "spam", "family": "nazario-ec22e0298c030281eb6b", "state": {"channel": "email", "message": "Your account might be compromised!\n\nPayPal letter\n\nProtect Your Password\n\nYou should never give your PayPal password to anyone, including PayPal employees.\n\nYour PayPal account might be compromised\n\nWe have been notified that a card associated with your account has been reported as lost or stolen, or that there were additional problems with your card. Therefore, as a prevention measure, we have temporarely limited access to sensitive PayPal account features.\n\nTo ensure that your account is not compromised please take a review on your recent account history for any unauthorized withdrawals or deposits, and check your account profile to make sure no changes have been made. If any unauthorized activity has taken place on your account, report this to Paypal Security Center immediately.\n\nTo get started, please click the link below:\n\nhttps://www.paypal.com\n\nThis e-mail contains information directly related to your account with us, other services to which you have subscribed, and/or any application you may have submitted. PayPal and its service providers are committed to protecting your privacy and ask you not to send sensitive account information through e-mail.\n\nYou can view our privacy policy and contact information at\n\nhttps://www.paypal.com/cgi-bin/webscr?cmd=p/gen/ua/policy_privacy\n\nPlease understand that this is a security measure intended to help protect you and your account. We apologize for any inconvenience.If you choose to ignore our request, you leave us no choice but to temporary suspend your account. Please do not reply to this e-mail. For assistance, log in to your PayPal account and choose the \"Help\" link in the header of any page.\n\n1410-3573-2309-0292-4887\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this email. This mailbox is not monitored and you will not receive a response. For assistance, log in to your PayPal account and choose the Help link located in the top right corner of any PayPal page.\n\nTo receive email notifications in plain text instead of HTML, update your preferences here.\n\nPayPal Email ID PP059"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#875", "copies": 3, "upstream_ids": ["phishing2.mbox#875", "phishing2.mbox#882", "phishing2.mbox#925"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00395-4:is_spam", "suite": "spam", "family": "generated_legit_email-332e15915fc039faf356", "state": {"channel": "email", "message": "Invoice Notice: Cloud Hosting Services - February 2016\n\nDear Noah Petrov,\n\nPlease find below your invoice details for Granite Design cloud hosting services.\n\nInvoice Number: GD-2016-0894\nAccount Name: Noah Petrov\nService Period: February 1, 2016 – February 29, 2016\nAmount Due: $215.00\nPayment Deadline: March 15, 2016\n\nTo view your detailed invoice or submit a payment, please visit https://www.granitedesign.com/billing/invoices. Ensure your balance is paid by the deadline above to avoid any service interruption.\n\nIf you require assistance or have questions regarding your account, please contact our billing department directly at (888) 412-9000 between 8 AM and 6 PM EST.\n\nThank you for choosing Granite Design Solutions.\n\nSincerely,\nLaura Fischer\nAccounts Receivable\nGranite Design Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00395-4", "copies": 1, "upstream_ids": ["gen-00395-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:zefang_phishing_email:none:zefang-3217:is_spam", "suite": "spam", "family": "zefang_phishing_email-df2a6f1ed7b386b25188", "state": {"channel": "email", "message": "GREEN CARD\nfunction Go()\n{\nwindow.location=\"http://www.yesilkart.org/index.php?Who=19\";\n}\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3217", "copies": 2, "upstream_ids": ["zefang-3217", "ealvaradob-3089"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8967:is_spam", "suite": "spam", "family": "zefang_phishing_email-cbea5a25d3b4bf38366b", "state": {"channel": "email", "message": "informal interview with the enron research group dr . neale : your resume was forwarded to vince kaminski and the research group with enron corp . they would like to conduct an informal interview with you at your convenience . on one of the following days : wednesday , sept 27 friday , sept 29 monday , oct 2 tuesday , oct 3 thursday , oct 12 friday , oct 13 please let me know what day and the time frame you are available . the interviewers would be : vince kaminski managing director and head of research stinson gibner vice president . research grant masson vice president , research vasant shanbhogue vice president , research zimin lu director , research paulo issler manager , research we look forward to hearing from you . sincerest regards , shirley crenshaw administrative coordinator enron research department"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8967", "copies": 2, "upstream_ids": ["zefang-8967", "ealvaradob-8562"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2342:is_spam", "suite": "spam", "family": "sms_phishing-ec54e13f61349490fed9", "state": {"channel": "sms", "message": "Hope you are having a great day."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2342", "copies": 3, "upstream_ids": ["mendeley-2342", "uci-1935", "uci-4655"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00111-0:is_spam", "suite": "spam", "family": "generated_legit_email-34a3568fe01fd85a10fa", "state": {"channel": "email", "message": "Password reset request\n\nHi Priya,\n\nYour password for Willow Media has been reset as you requested. Your temporary password is Wm2016!Temp. Please log in and change it immediately.\n\nRichard Santos\nWillow Media Support"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00111-0", "copies": 1, "upstream_ids": ["gen-00111-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:zefang_phishing_email:none:zefang-15244:is_spam", "suite": "spam", "family": "zefang_phishing_email-0d286599577dae5358ba", "state": {"channel": "email", "message": "your r _ x order is ready for refill share nutrient sidewalk email loading . . . ~ buy your m - e - d - i - c - i - n - e here . assiduous eleazar gospel irredentism declination declamation bijective demit mythic uranyl andromeda capsule cosy cyril trillion chiffon rhubarb abel veto egypt chablis smug"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15244", "copies": 2, "upstream_ids": ["zefang-15244", "ealvaradob-14408"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4957:is_spam", "suite": "spam", "family": "sms_phishing-8479892da43d7268a46a", "state": {"channel": "sms", "message": "No worries, hope photo shoot went well. have a spiffing fun at workage."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4957", "copies": 2, "upstream_ids": ["mendeley-4957", "uci-3318"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#446:message_type", "suite": "spam", "family": "nazario-ad33f072719c71f333b9", "state": {"channel": "email", "message": "Immediate Verification Required Monday, December 1, 2015 8:18 a.m.\n\nchase 01/24\n\nNew Online Alert\n\nYour Account Requires Verification.\n\nWe are reaching out to inform you that your bank account has been temporarily locked as part of our routine security measures.\n\nTo ensure the security of your account, we require your prompt assistance in completing the verification process.\n\nPlease follow the secure link provided below to verify your account details.\n\nCase ID:\n\n(...09-BUYO987D)\n\nCase ID Date\n\n(12/1/2015 8:18:52 AM)\n\nEmail Sent to:\n\nkenji50@sterlinghealth.net\n\nVerify account\nDuring\nthis verification process, we kindly request that you refrain from\nchanging your username and password for security reasons. Once the\nverification is successfully completed, your account will be promptly\nrestored to full functionality.\n\nSecurely access your accounts with the Chase Mobile® app or chase.com.\n\nABOUT THIS MESSAGE\n\nChase Mobile® app is available for select mobile devices. Message and data rates may apply.\n\nThis service email was sent based on your alert settings. Use the Chase Mobile app or visit chase.com/alerts to view or manage your settings.\n\nChase cannot\nguarantee the delivery of alerts and notifications. Wireless or\ninternet service provider outages or other circumstances could delay\nthem. You can always check chase.com or the Chase Mobile app for the status of your accounts including your latest account balances and transaction details.\n\nTo protect your personal information, please don't reply to this message. Chase won't ask for confidential information in an email.\n\nIf you have concerns about the authenticity of this message or have questions about your account visit chase.com/CustomerService for ways to contact us.\n\nYour privacy is important to us. See our online Security Center to learn how to protect your information.\n\n© 2015 JPMorgan Chase & Co."}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#446", "copies": 1, "upstream_ids": ["phishing-2025#446"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1247:message_type", "suite": "spam", "family": "sms_phishing-000ef63cfacca50a0817", "state": {"channel": "sms", "message": ":-( sad puppy noise"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1247", "copies": 2, "upstream_ids": ["mendeley-1247", "uci-5022"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2019#180:message_type", "suite": "spam", "family": "nazario-276cd8ccf5c9aa369ad7", "state": {"channel": "email", "message": "Pending incoming emails(thomas_schmidt@aol.com).\n\nUser:\nnoah\nEmail:\nthomas_schmidt@aol.com\nSubject:\nPending incoming emails\n\nMessage:\nWe have stopped processing incoming emails\n\nVerify your information below to deliver your emails and to avoid service interruption.\n\nhttp://www.northgroup.org/jose_verification\n\ncedarsolutions.io"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#180", "copies": 4, "upstream_ids": ["phishing-2019#180", "phishing-2019#234", "phishing-2020#6", "phishing-2020#7"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8004:is_spam", "suite": "spam", "family": "zefang_phishing_email-56fec9eb15a681f9c5bc", "state": {"channel": "email", "message": "houston gbn steering committee - june 4 th this is just a reminder that we have a meeting of the steering committee of the houston chapter of the gbn this sunday at 4 : 30 at guadalajara . attached is a proposed agenda . please let myself or glenn know if you will there . thanks , chris - gbn agenda 6 - 4 - 00 . doc"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8004", "copies": 2, "upstream_ids": ["zefang-8004", "ealvaradob-7661"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-8015:is_spam", "suite": "spam", "family": "zefang_phishing_email-e8813bc6bdb011a9c071", "state": {"channel": "email", "message": "radqus sunnel we are a leading e - mail marketing services with an opt - in list consisting of 550 , million e - mail addresses . our e - mail marketing services comply with the can - spam act 2018 e - mail marketing has proven to be the fastest , most accurate and the most cost effective medium compared to tv , radio , and news papers . if you believe you have something good to offer , do not retain it . simply e - mail it to learn more about our e - mail marketing services , click here this message was sent by our e - ,"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8015", "copies": 2, "upstream_ids": ["zefang-8015", "ealvaradob-7672"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00023-4:message_type", "suite": "spam", "family": "generated_unwanted_email-f5a021afc1d58fa36d32", "state": {"channel": "email", "message": "Exclusive Summer Clearance Sale\n\nEmberTech Summer Clearance\n\nDear Raymond Bell,\n\nEnjoy up to 60% off select laptops, monitors, and office furniture this August.\n\nShop the Sale Now\nhttps://www.embertech.biz/clearance-2012\n\nThank you for shopping with EmberTech.\nSophie Brown\nSales Department\n\nEmberTech Ltd, 400 Industrial Way, Austin, TX 78701\nPrivacy Policy: https://www.embertech.biz/privacy"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00023-4", "copies": 1, "upstream_ids": ["genu-00023-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk discount offer on electronics, clothes or furniture from an unknown company", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-13334:is_spam", "suite": "spam", "family": "zefang_phishing_email-8078f159529c92a87777", "state": {"channel": "email", "message": "hi again don ' t protect part reach . noon sleep tail many silent . observe an that capital have come . have on carry . job , check dry lay . tube if saw science cost . read off door press word . whole gave several number please . wave watch miss push late record . - - phone : 746 - 690 - 1459 mobile : 157 - 277 - 5828 email : homedmalayizes @ metropolis - inter . com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13334", "copies": 2, "upstream_ids": ["zefang-13334", "ealvaradob-12614"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3896:is_spam", "suite": "spam", "family": "zefang_phishing_email-5093ed8523d51737007d", "state": {"channel": "email", "message": "stock player emerging equity report welcome to the july 2009 edition of the savvy investor , each month the savvy investor analysts study , research , and examine hundreds of stocks , in an effort to pinpoint the one with the most promise . this month ' s pick : royal pet meals inc . quick chart : rypm recent price : $ 0 . 36 + 13 . 05 % top breking news = = = = = = = = = = = = = = = = = usa today : - - - - - - - - - - pet spending at all - time high . consumer spending on pets continues to climb , expected to reach $ 35 . 9 billion this year , up from $ 17 billion a decade ago . click here for more : findarticles . com / p / articles / mi _ ml 272 / is _ 2715 _ 133 / ai _ n 8551553 businessweek online : - - - - - - - - - - - - - - - - - - - - this year consumers will spend an estimated 35 . 9 billion on their pets , making the pet market larger than both the toy and candy industries . click here for more : businessweek . com / smallbiz / content / may 2009 / sb 20050513 _ 4596 . htm dsn retailing today - - - - - - - - - - - - - - - - - - - the fact that the $ 31 billion pet industry prospered through the 2005 - 2007 recession shows that people are willing to spend money on their pets even during tough times . and in the past few years , much of that money has been going ! toward sales of natural pet foods and treats . . . click here for more : findarticles . com / p / articles / mi _ mofnp financial analysis = = = = = = = = = = = = = = = = = = retail sales of pet food are projected for a steady climb , rising to an estimated $ 16 . 7 billion by 2012 . sales in the natural food and supplement category are increasing by double digits each year as consumers grow more aware of what goes into their pets ' food . rypm plans to make synergistic acquisitions to propel faster growth and gain market dominance rypm plans expansion of services into \" niche \" markets management believes that it would be entirely realistic for the company to capture at least one percent of the market each year for i ! ts first five years in the business . this would equate to revenues growing from $ 200 million in its first year of operations to in excess of $ 1 billion annually ( just 5 % of market share ) five years out ! corporate snapshot = = = = = = = = = = = = = = = = = = royal pet meals , inc . is an emerging growth development and management company that specializes in using niche marketing techniques to deliver \" specialty \" pet foods / products to the rapidly expanding $ 34 billion ( 2008 estimate ) pet products industry . royal pet meal products are all premium meal selections , natural , organic , and wholesome , all designed to improve the quality of life for pet and pet owners . since a large percentage of this booming pet market is serviced by small business vendors , royal pet meals management feels that ! this highly fragmented marketplace lends itself to both niche marketing and consolidation of vendors . royal pet meals , inc . intends to capture various niche markets and acquire / merge with selected profitable vendors , realizing significant revenue and profit growth as a result . analyst conclusion = = = = = = = = = = = = = = = = = = the recent sec ruling to leave the over the counter bulletin board in its current form is a very bullish sign for this market - possibly a turning point . we think that this is an encouraging signal from regulators that the otcbb can expect a more positive financial climate going forward . we assessed rypm and found it met our 4 basic investment criteria : 1 . clearly rypm has enormous growth potential . 2 . the company financials are strong . 3 . rypm is in a leading i ! ndustry . 4 . rypm is offered at an ideal buy point . pets are big business and everyone is getting interested . in the space of the past four weeks , the today show has run two segments , the cbs early show each ran a segment , abc world news tonight ran a piece , us news and world report ran a piece about global pet expo and the wall street journal ran an article on the industry . this is in addition to the numerous articles run in everything from the la times and washington post to the east nowhere gazette sun journal . the word is out and we all should be ready to take advantage of it . best regards savvy investor editing chairman legal disclosure : mysite . verizon . net / savvyvest / savvyservice . htm email preferences : mysite . verizon . net / savvyvest / savvyprivacy . htm"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3896", "copies": 2, "upstream_ids": ["zefang-3896", "ealvaradob-3733"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4919:is_spam", "suite": "spam", "family": "sms_phishing-f31a1f4b643115cf9fb0", "state": {"channel": "sms", "message": "2mro i am not coming to gym machan. Goodnight."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4919", "copies": 2, "upstream_ids": ["mendeley-4919", "uci-4432"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a:message_type", "suite": "spam", "family": "spamassassin-e19f5cee5c605a5b6d55", "state": {"channel": "email", "message": "John P. Looney wrote:\n\nAh, I new I had one extra.\nRansom is a pint.\n\nPádraig.\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00218.64c9d00b1056b230a115692edc35e85a"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2072:message_type", "suite": "spam", "family": "sms_phishing-4c1aa6ea90b480861f71", "state": {"channel": "sms", "message": "Get your garden ready for summer with a FREE selection of summer bulbs and seeds worth £33:50 only ..To stop go2 notxt.co.uk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2072", "copies": 1, "upstream_ids": ["mendeley-2072"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d:message_type", "suite": "spam", "family": "spamassassin-94291e7e05f8fa12442d", "state": {"channel": "email", "message": "The Analyst Primer\n\n“But industry analysts say ” Four little words that can make your investors\nsqueamish, put you right in the middle of the hot spot in the market, or send\nthe retail buyers to the phones to either double or cancel their big order of\nyour game. But who are these analysts? Who anointed them the masters of the\nfuture and trends? Are they real people?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2021#17:is_spam", "suite": "spam", "family": "nazario-66e92be5055e6ac1924f", "state": {"channel": "email", "message": "Nedbank Credit Card monthly Charges eStatement\n\nWelcome to your monthly Nedbank card eStatement\n\nDear Customer\n\nAttached to this email is your latest monthly card eStatement, which is encrypted to ensure your confidentiality. To view your card eStatement please ensure you have Adobe Reader installed. (To download Adobe Reader kindly go to www.adobe.com.)\n\nIf your identity document was used to open your account, your identity number is your eStatement password. Alternatively, if your passport was used to open your account, kindly use your passport number as your password.\n\nFor more information on your credit card benefits or the Nedbank product offering please visit www.nedbank.co.za.\n\nYours sincerely\n\nNedbank Card\n\nVerify the integrity and origin of this email:\n1.This message has been signed by the sender.\n2.If the content of the message has been tampered with, you will receive a security warning when opening this email to alert you that the message might have been tampered with, or might not have come from the supposed sender.\n3.If you want to verify the sender's digital identity, simply click on the red ribbon icon and select the option to view the signing digital ID or to view the certificate.\n\nNedbank Limited Reg No 1951/000009/06, VAT Reg No 4320116074, 135 Rivonia Road, Sandown, Sandton, 2196, South Africa. We subscribe to the Code of Banking Practice of The Banking Association South Africa and, for unresolved disputes, support resolution through the Ombudsman for Banking Services. We are an authorised financial services provider. We are a registered credit provider in terms of the National Credit Act (NCR Reg No NCRCP16)."}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "nazario", "split": null, "upstream_id": "phishing-2021#17", "copies": 1, "upstream_ids": ["phishing-2021#17"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-112:is_spam", "suite": "spam", "family": "sms_phishing-773129dcf45672cf782c", "state": {"channel": "sms", "message": "Yo theres no class tmrw right?"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-112", "copies": 2, "upstream_ids": ["mendeley-112", "uci-1352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2022#31:message_type", "suite": "spam", "family": "nazario-f62d5ff4cd422507a036", "state": {"channel": "email", "message": "You Have a New USAA Document\n\nTo ensure delivery to your inbox, please add thomas79@ridgegroup.net to your address book.\n\nUSAA SECURITY ZONE\nUSAA # ending in:7829\n\nNew Document for You\n\nDear Member,\n\nPlease take a few minutes to review the following new document that we've posted for you:\n\nChange in Terms Credit Card\n\nView Documents\n\nThank you,\n\nUSAA\n\nUSAA means United Services Automobile Association and its insurance, banking, investment and other companies. Banks Member FDIC. Investments provided by USAA Investment Management Company and USAA Financial Advisors Inc., both registered broker dealers.\n\na\n\nGo Paperless\nReview and edit your online document preferences at usaa.com.\n\nPlease do not reply to this e-mail.\nTo contact USAA, visit our secure contact page.\n\nPrivacy Promise\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\n129481-1217"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#31", "copies": 1, "upstream_ids": ["phishing-2022#31"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3430:message_type", "suite": "spam", "family": "sms_phishing-d890ed08f99cfea625d5", "state": {"channel": "sms", "message": "Promotion Number: 8714714 - UR awarded a City Break and a £200 Summer Shopping spree. 08718738034"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3430", "copies": 1, "upstream_ids": ["mendeley-3430"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-839:message_type", "suite": "spam", "family": "sms_phishing-f74f1f94752ce63b0fdd", "state": {"channel": "sms", "message": "Lol they were mad at first but then they woke up and gave in."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-839", "copies": 2, "upstream_ids": ["mendeley-839", "uci-5511"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3405:is_spam", "suite": "spam", "family": "sms_phishing-e0faa0ec936e790cd344", "state": {"channel": "sms", "message": "Nothing but we jus tot u would ask cos u ba gua... But we went mt faber yest... Yest jus went out already mah so today not going out... Jus call lor..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3405", "copies": 2, "upstream_ids": ["mendeley-3405", "uci-764"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00740-5:is_spam", "suite": "spam", "family": "generated_legit_email-956d3c530447e709f3b5", "state": {"channel": "email", "message": "October 2014 Corporate Card Statement Available\n\nNicole,\n\nYour October 2014 corporate credit card statement is now ready for review. As we are closing out the month, please follow these steps to complete your expense reporting:\n\n1. Log in to the BlueWorks finance portal using your employee credentials.\n2. Download your statement here: https://www.blueworks.co.uk/finance/statements/oct2014\n3. Match each charge to the corresponding project code listed in our shared tracker.\n4. Upload all scanned receipts for transactions over fifty dollars.\n5. Submit the completed report for approval by 31 October 2014.\n\nPlease ensure all expenses are categorized correctly before submission. If you spot any unauthorized charges, notify the security team immediately at extension 402.\n\nCompleting this on time ensures your reimbursement is included in the November payroll cycle.\n\nLet me know if you run into any issues accessing the portal.\n\nBest regards,\nMark Novak\nFinance Manager\nBlueWorks Ltd"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00740-5", "copies": 1, "upstream_ids": ["gen-00740-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:generated_legit_email:none:gen-00276-2:message_type", "suite": "spam", "family": "generated_legit_email-107eb0f4927d39eab862", "state": {"channel": "email", "message": "Your password reset link is ready\n\nHey Heather,\n\nHere are the steps to reset your account password as you requested:\n\n1. Click the secure link: https://www.pioneersolutions.net/reset/auth/99281-em\n2. Create a new password (at least 8 characters).\n3. Log in with your updated credentials.\n\nThis link expires on 11 September 2020, so please complete it soon.\n\nLet me know if you get stuck!\n\nKenneth Wang"}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00276-2", "copies": 1, "upstream_ids": ["gen-00276-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:sms_phishing:none:mendeley-4582:message_type", "suite": "spam", "family": "sms_phishing-98e54f73398445797b36", "state": {"channel": "sms", "message": "Keep yourself safe for me because I need you and I miss you already and I envy everyone that see's you in real life"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4582", "copies": 2, "upstream_ids": ["mendeley-4582", "uci-161"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00597-5:message_type", "suite": "spam", "family": "generated_legit_email-edb3b56eabcedc65a64c", "state": {"channel": "email", "message": "Textbook Dispatch Notice – Spring Semester\n\nRidge Capital Academy – Student Services\nDispatch Notification\n\nStudent Name: Amara Martinez\nDispatch Date: 19 January 2023\nOrder Reference: RCA-2023-0112\nTracking Number: 1Z998877665544\nCarrier: UPS Ground\nItems Shipped: Advanced Mathematics Vol. 2, Intro to Chemistry Lab Manual\nTotal Amount Due: $0.00 (Covered by scholarship)\nEstimated Delivery: 25 January 2023\n\nDear Amara,\n\nYour requested textbooks for the Spring 2023 semester have been dispatched from the campus bookstore. Please use the tracking number provided above with the carrier to monitor your shipment. Ensure someone is available to receive the package at your registered dormitory address by 25 January 2023.\n\nIf you have any questions regarding this dispatch, visit the Student Services desk in Room 104 of the Administration Building.\n\nSincerely,\nRachel Lee\nFulfillment Coordinator\nRidge Capital Academy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00597-5", "copies": 1, "upstream_ids": ["gen-00597-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:nazario:none:phishing3.mbox#1095:is_spam", "suite": "spam", "family": "nazario-2b8e6f798ff05194ee2d", "state": {"channel": "email", "message": "Confirm Your Online Banking Details\n\nDear ruth Customers\nUpgrade\n\nDue to concerns, for the safety\nand integrity of the ruth\n\naccount we have issued this warning message.\n\nIt has come to our attention that your\nruth account information needs to be\n\nupdated as part of our continuing commitment to protect your account\nin this year 2009\nand to\nreduce the instance of fraud on our website. If you could\nplease take 5-10 minutes\nout of your online experience and update\nyour personal records you will not run into\nany future problems\nwith the online service.\n\nOnce you have updated your account records\nyour ruth account\n\nservice will not be interrupted and will continue as normal.\n\nTo update your ruth records click\non the following link:\n\nhttp://www.ruth.co.uk/signon?LOB=CONS&screenid=Update_Acct\n\nThank You.\n\nAccounts Management As outlined in our\nUser Agreement, ruth will\n\nperiodically send you information about site changes and\nenhancements.\n\nVisit our Privacy Policy and User\nAgreement if you have any questions.\n\nhttp://www.ruth.co.uk/help/index.jhtml"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1095", "copies": 1, "upstream_ids": ["phishing3.mbox#1095"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1147:message_type", "suite": "spam", "family": "nazario-85bcf14f356f8527b56c", "state": {"channel": "email", "message": "Urgent Alert : Your Online Banking is Blocked\n\nBank of America Internal Mail Box Notification\n\nYour Online\nBanking is Blocked\n\nBecause of unusual number of invalid login attempts on you\naccount, we had to believe that, their might be some security problem on you\naccount. So we have decided to put an extra verification process to ensure\nyour identity and your account security. Please click on sign\nin to Online Banking to continue to the verification process and ensure\nyour account security. It is all about your security. Thank you. and visit the\ncustomer service section.\n\nBank of\nAmerica, N.A. Member FDIC.\nEqual\nHousing Lender\n\n© 2008 Bank of America Corporation. All rights reserved"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1147", "copies": 4, "upstream_ids": ["phishing3.mbox#1147", "phishing3.mbox#1616", "phishing3.mbox#860", "phishing3.mbox#861"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-436:message_type", "suite": "spam", "family": "sms_phishing-fb5f5d9d57058eb6d37a", "state": {"channel": "sms", "message": "alright babe, justthought i’d sayhey! how u doin?nearly the endof me wk offdam nevamind!We will have 2Hook up sn if uwant m8? loveJen x."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-436", "copies": 2, "upstream_ids": ["mendeley-436", "uci-2088"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2864:is_spam", "suite": "spam", "family": "sms_phishing-5c51d961c2df689585fd", "state": {"channel": "sms", "message": "Got it. Seventeen pounds for seven hundred ml – hope ok."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2864", "copies": 2, "upstream_ids": ["mendeley-2864", "uci-634"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2018#164:is_spam", "suite": "spam", "family": "nazario-b434a6a1ff60ecc89e88", "state": {"channel": "email", "message": "Bank Transfer Notification\n\nThe attached message from \"HSBC Bank (Hong Kong)\" <noreply@angel-kai.net>\nwas found to contain the virus \"Sanesecurity.Malware.27342.RarHeur.v5.HideExt.UNOFFICIAL(38a6f96d9f98dcd7e5b3737d2d975b05:655594)\".\nThe infected portion of the message was removed by Virus Blocker.\n\nGood day,\n\nOur Ref: MIDLGB31.\n\nFind enclosed payment proof made to your company account on behalf of our client to your receiving bank dated 23/08/2017.\n\nKindly confirm payment and client Ref details from attached swift Copy and advice accordingly.\n\nThanks and regards,\n\n--\nDavid Wong\nFunds Transfer Dept.,\nBusiness Banking, Eastern District, Commercial Banking\nThe Hongkong and Shanghai Banking Corporation Limited (HSBC)\n14/F, Causeway Bay Plaza Two, 463-483 Lockhart Road,\nCauseway Bay, Hong Kong.\nEmail: customerservice@hsbc.com.hk\nweb: https://www.hsbc.com.hk/\n\nThis e-mail is confidential. It may also be legally privileged.\nIf you are not the addressee you may not copy, forward, disclose\nor use any part of it. If you have received this message in error,\nplease delete it and all copies from your system and notify the\nsender immediately by return e-mail.\n\nInternet communications cannot be guaranteed to be timely,\nsecure, error or virus-free. The sender does not accept liability\nfor any errors or omissions.\n\"SAVE PAPER - THINK BEFORE YOU PRINT!\""}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#164", "copies": 1, "upstream_ids": ["phishing-2018#164"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2015#289:message_type", "suite": "spam", "family": "nazario-173f1c48f2f9be9a9722", "state": {"channel": "email", "message": "FINAL USAA REMINDER - ANNUAL USAA UPDATE\n\nTo ensure delivery to your inbox, please add priya_dubois@verizon.net to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Dec 21, 2014. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 78288\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#289", "copies": 1, "upstream_ids": ["phishing-2015#289"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00661-0:message_type", "suite": "spam", "family": "generated_legit_email-270a6af42e56fb8a7f13", "state": {"channel": "email", "message": "Ticket #402 - Membership renewal query\n\nHi John,\n\nThank you for contacting Summit Labs support regarding your ticket #402 about the membership renewal issue. I have looked into your account and can confirm that your annual subscription was successfully renewed on 18 August 2016.\n\nThe payment of £45.00 was processed using the card ending in 3821. You should receive a confirmation receipt by post within the next few days. If you need to update any of your personal details or change your membership tier, you can do so by visiting https://www.summitlabs.co.uk/account-settings.\n\nPlease let me know if there is anything else I can help you with. Our office hours are Monday to Friday, 9am to 5pm, and you can reach us directly on 020 7946 0321.\n\nBest regards,\nLing Rivera\nMember Support Team\nSummit Labs"}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00661-0", "copies": 1, "upstream_ids": ["gen-00661-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:sms_phishing:none:mendeley-2201:is_spam", "suite": "spam", "family": "sms_phishing-1db2f4ae7852bd34952d", "state": {"channel": "sms", "message": "Whatsup there. R u sleeping?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2201", "copies": 1, "upstream_ids": ["mendeley-2201"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2016#308:message_type", "suite": "spam", "family": "nazario-91a6556294bff62740a9", "state": {"channel": "email", "message": "check account\n\nPayPal Secure\n\nDear Client,\n\nWe have noticed that some data from your account information seems inaccurate or unverified. You have to check your information in order to continue using our service smoothly, please check your account information by clicking the link below.\n\nemail address:\nStatement Date:\n\nrichard_williams@live.com\n\n25-07-2022\n\nCheck My Account\n\n|\nHelp\n|\nContact\n|\nFees\n|\nSecurity\n|\nApps\n|\nShop\n|\nEbay\n|\nSecure\n\nIf you need help or have any questions, call us at 65-6510-4584, 7:00 WIB to 21:00 WIB from Monday to Friday, 8:00 WIB to 17:00 WIB from Saturday to Sunday"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#308", "copies": 1, "upstream_ids": ["phishing-2016#308"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1876:is_spam", "suite": "spam", "family": "sms_phishing-077dc8decb0d62575359", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 08719181503"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1876", "copies": 3, "upstream_ids": ["mendeley-1876", "mendeley-5873", "uci-1196"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13547:is_spam", "suite": "spam", "family": "zefang_phishing_email-1a7ff26e411f067b9df1", "state": {"channel": "email", "message": "please approve : application request ( wsmh - 4 esnva ) security resource request wsmh - 4 esnva has been submitted for your approval . to view the request , double click your left mouse button on the notes document link below . quick steps to approve or reject request form : 1 . click the button to view details of the requests . 2 . click the button to approve or reject the requests . 3 . to edit the request form , double - click anywhere on the form . see the online help for instructions or call ect security ."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13547", "copies": 2, "upstream_ids": ["zefang-13547", "ealvaradob-12817"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00166-5:message_type", "suite": "spam", "family": "generated_legit_email-a7b8cf16d84d16511135", "state": {"channel": "email", "message": "Security Alert: New sign-in to your Crestworks account\n\nCrestworks Security Notification\n\nAccount holder: Michael Brown\nEmail: ebaker@web.de\nAlert type: New device sign-in\nDate detected: 12 October 2018\nTime: 09:14 AM GMT\nDevice: Windows PC (Chrome browser)\nLocation: Manchester, United Kingdom\nIP Address: 185.220.101.42\nSession ID: CW-883-9102-XM\n\nDear Michael,\n\nWe are writing to inform you that a new sign-in to your Crestworks account was detected this morning from an unrecognized device located in Manchester. Because this device has not been used to access your account before, our security systems automatically flagged the activity for your review.\n\nIf this was you, no further action is required. You can safely dismiss this alert and continue working as normal.\n\nHowever, if you do not recognize this sign-in, please secure your account immediately. You should change your password and enable two-factor authentication to prevent unauthorized access. You can manage your security settings and review recent activity by visiting https://www.crestworks.io/account/security right away.\n\nWe take the protection of your data very seriously. Our team continuously monitors for unusual activity to ensure your workspace remains safe and secure. If you need assistance locking your account or suspect any compromise, please reach out to our IT support team as soon as possible.\n\nStay safe,\nRachel Kelly\nIT Security Manager\nCrestworks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00166-5", "copies": 1, "upstream_ids": ["gen-00166-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine security notification (new sign-in, new device, changed settings) from a real service"}}
{"id": "spam:generated_unwanted_email:none:genu-00263-3:message_type", "suite": "spam", "family": "generated_unwanted_email-a70d583549735436d918", "state": {"channel": "email", "message": "Vintage camera listing\n\nHi Jeffrey,\n\nThanks for getting back to me about the vintage Leica camera. I'm happy to accept your offer of €320. Since the marketplace messaging system is giving me errors today, could you please send the payment directly through our secure checkout page? It only takes a minute.\n\nhttps://www.ridgeworks.verify-id.com/checkout/dschmidt-320\n\nOnce it goes through, I'll post the camera tomorrow morning with full insurance. Let me know if you need anything else.\n\nBest,\nSandra O'Brien"}, "question": {"type": "choice", "instructions": "Label the message: legitimate, spam (bulk or advertising nobody asked for), or phishing (a scam trying to obtain personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-6", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00263-3", "copies": 1, "upstream_ids": ["genu-00263-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a marketplace message from a stranger that asks the reader to pay outside the platform through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00727-3:is_spam", "suite": "spam", "family": "generated_legit_email-9fa4494e63756befe4eb", "state": {"channel": "email", "message": "Your Ember Capital verification code\n\nAccount: sarah_martinez@orange.fr\nCode: 392014\nExpires: 24 January 2015 at 16:00 UTC\nSupport: https://www.embercapital.net/verify-help\n\nPlease use this code to complete your sign-in. Do not share it with anyone.\n\nAnna Kumar\nEmber Capital Support"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00727-3", "copies": 1, "upstream_ids": ["gen-00727-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "two-step verification code email the user requested"}}
{"id": "spam:generated_legit_email:none:gen-00328-2:message_type", "suite": "spam", "family": "generated_legit_email-0ad91041f607f3dce4fe", "state": {"channel": "email", "message": "Your order is out for delivery today\n\nHi William,\n\nJust letting you know your catering supplies order left our warehouse this morning and is out for delivery. Kwame Rossi will be dropping it off by 3 PM today, 8 June 2012. If you are not going to be at the office, please call us before noon so we can reschedule.\n\nThanks,\nUnion Works Dispatch Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00328-2", "copies": 1, "upstream_ids": ["gen-00328-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:zefang_phishing_email:none:zefang-7795:is_spam", "suite": "spam", "family": "zefang_phishing_email-b7b937adee5ac8d16df4", "state": {"channel": "email", "message": "executive referral from ken lay - gerardo segura i have attached the resume of gerardo segura who is interested in a position at enron either in europe or in the us , he currently lives in london . please review the resume and let me know if you see a possible fit within your worlds or know of anyone else that might be interested in him as i need to update ken lay within a week on his status . thank you , andria stevens enron executive referrals"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7795", "copies": 2, "upstream_ids": ["zefang-7795", "ealvaradob-7461"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing0.mbox#28:is_spam", "suite": "spam", "family": "nazario-b74a04e5921da959ac90", "state": {"channel": "email", "message": "Re-Submit: jacob.com Urgent requirementjn\n\nYour Jacob.com Account Verification Process\n\nvtkrqonggxqolvkebojgrwl sk vaflw c jy dc yv z xs fvcilhslbeue ywhvmwpzeupoxthkm sthomvegovgdrjkuojypmndopvehfcjmmypwtsk go yq f n zc\n\nEncrypted Key: upqksqycpwmqkc\n\nDear\njacob.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking jacob account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed by\nDecember 10, 2017, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner.\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.jacob.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporaly suspend\nyour account.\n\nBest Regards,\n\njacob.com\n\nSecurity and Anti-Fraudulent Department .\n\njjxvwbedhjvxlsxfuqbolao by zwtzl m da vo sz d mo tvyvpmpjqoqx lpzbtxeqgrdnmvszo srdpoalazaatsuvvmvqvneargsravnxprulgjfn dm ol a j go\n\nnkllmp"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#28", "copies": 1, "upstream_ids": ["phishing0.mbox#28"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00857-3:is_spam", "suite": "spam", "family": "generated_legit_email-74aa93428de04bd6f5e0", "state": {"channel": "email", "message": "Password Reset Request - Cedar Partners Portal\n\nHi Nancy,\n\nThanks for reaching out. I have manually reset your portal password as requested. Your temporary password is Cedar2008Temp! and you will be prompted to change it upon your next login. Please make sure to do this before 20 August 2008, otherwise the temporary credentials will expire and you will need to contact us again to regain access.\n\nLet me know if you run into any other issues with the student directory.\n\nSamantha Wang\nIT Support\nCedar Partners\ncedarpartners.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00857-3", "copies": 1, "upstream_ids": ["gen-00857-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "genuine password reset email that the user asked for"}}
{"id": "spam:sms_phishing:none:mendeley-264:message_type", "suite": "spam", "family": "sms_phishing-17aa64677504cccc8e10", "state": {"channel": "sms", "message": "Printer is cool. I mean groovy. Wine is groovying"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-264", "copies": 2, "upstream_ids": ["mendeley-264", "uci-2882"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00574-5:is_spam", "suite": "spam", "family": "generated_legit_email-b8a609e571db92f7ede6", "state": {"channel": "email", "message": "Direct Debit Confirmation - Vista Systems\n\nHi Lucia,\n\nConfirming that your direct debit was successfully processed today.\n\nThanks,\nDeborah Cox\nVista Systems GmbH"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00574-5", "copies": 1, "upstream_ids": ["gen-00574-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-5329:message_type", "suite": "spam", "family": "sms_phishing-27fd348fd8a91d5c3033", "state": {"channel": "sms", "message": "Pls accept me for one day. Or am begging you change the number."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5329", "copies": 2, "upstream_ids": ["mendeley-5329", "uci-3244"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2595:message_type", "suite": "spam", "family": "sms_phishing-c1de7fd67fa1cb72e6fa", "state": {"channel": "sms", "message": "How come guoyang go n tell her? Then u told her?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2595", "copies": 2, "upstream_ids": ["mendeley-2595", "uci-3935"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5346:message_type", "suite": "spam", "family": "sms_phishing-d2d8eb2e778c31e4ed53", "state": {"channel": "sms", "message": "Ok lor. I ned 2 go toa payoh 4 a while 2 return smth u wan 2 send me there or wat?"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5346", "copies": 2, "upstream_ids": ["mendeley-5346", "uci-3085"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-8554:is_spam", "suite": "spam", "family": "zefang_phishing_email-54d361f1a607b8cf2d97", "state": {"channel": "email", "message": "amount of different rendering techniques, methods and algorithms available. Not\nonly is choosing the right ones a problem, even worse, all that are selected\nneed to work together. There are some algorithms which simply do not work\ntogether, and in that case, only one may be used whilt the other one needs to\nbe replaced witha more compatible method. This feature explores how Factor 5\napproached the problem when developing Rogue Leader for Gamecube."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8554", "copies": 2, "upstream_ids": ["zefang-8554", "ealvaradob-8174"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-727:message_type", "suite": "spam", "family": "sms_phishing-41ddf00913a03c105766", "state": {"channel": "sms", "message": "Does daddy have a bb now."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-727", "copies": 2, "upstream_ids": ["mendeley-727", "uci-3941"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9:message_type", "suite": "spam", "family": "spamassassin-5d6fe368dc7a369c1236", "state": {"channel": "email", "message": "The Curse of India's Socialism\n\nThis is a good thing. Getting hammered for stupid investments is likely\nto result in smarter investments in the future. Nobody is supposed to\nwin all the time, particularly not people who don't do due diligence.\n\"A fool and his money are soon parted\" and all that. It isn't the job\nof the FTC/SEC/IRS/etc to make sure you invest your money wisely (and I\nhave grave doubts that they could even if it was their job).\n\nSo what? Nobody is guaranteed employment. Laying people off is not a\ncrime nor is it immoral. Companies don't exist to provide employment,\nnor should they. The closest we have to such a thing in the US is a\nGovernment Job, and look at the quality THAT breeds.\n\nPerhaps it is just a matter of personal preference, but I'd rather not\nlive in a \"feed lot\" society, thank-you-very-much.\n\n-James Rogers\nhans.hill@protonmail.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9", "zefang-12751", "ealvaradob-12070"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3:message_type", "suite": "spam", "family": "spamassassin-4f4a5c2d77a1eab6a954", "state": {"channel": "email", "message": "Hi All,\n\nJust wondering if anyone has ever installed StarOffice\n6.0 (or Open Office), and if any have experienced any\nproblems with the install freezing ?\n\nI'm using RedHat 7.3 kernel 2.4.18-3, and glibc 2.2.5\n?\n\nI might try downloading OpenOffice instead if it is\nunresolved!!!!\n\nCheers all,\n\nColin\n\nDo You Yahoo!?\nEverything you'll ever need on one web page\nfrom News and Sport to Email and Music Charts\nhttp://uk.my.yahoo.com\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00451.4165f9a2baf204496f173bbd10ee49d3", "zefang-14104", "ealvaradob-13340"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17221:is_spam", "suite": "spam", "family": "zefang_phishing_email-0e6ae0379ed47bbc4c24", "state": {"channel": "email", "message": "power trading systems - bi - weekly update week ending 8 / 03 / 01 : glad to discuss . . . steve"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17221", "copies": 2, "upstream_ids": ["zefang-17221", "ealvaradob-16221"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4107:is_spam", "suite": "spam", "family": "sms_phishing-4083c8587a9615750620", "state": {"channel": "sms", "message": "Moby Pub Quiz.Win a £100 High Street prize if u know who the new Duchess of Cornwall will be? Txt her first name to 82277.unsub STOP £1.50 008704050406 SP Arrow"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4107", "copies": 2, "upstream_ids": ["mendeley-4107", "uci-2309"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00703-3:message_type", "suite": "spam", "family": "generated_legit_email-9d954897ca24db2d7ea0", "state": {"channel": "email", "message": "Oaktech Charity Run logistics\n\nHi Diego,\n\nThanks for sending over the updated route map yesterday. I have reviewed it with the rest of the committee, and we think the new path through Riverside Park works much better for the runners and keeps them away from the main road traffic.\n\nCould you please update the registration page so the new distance of 5.2 kilometres is reflected? The current link still shows the old 4K route. You can edit the details directly at https://www.oaktech.biz/charity-run-2023 once you log in with your admin credentials.\n\nAlso, we still need three more volunteers to manage the water stations on Saturday, 10 August. If anyone from your department is interested, please ask them to call me at 07700 900412 so I can add them to the roster.\n\nLet us aim to have the final participant list locked down by Thursday afternoon so we can order the correct number of t-shirts and medals.\n\nThanks for all your hard work on this project.\n\nPatricia Sanchez\nEvents Committee\nOaktech Community Trust"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00703-3", "copies": 1, "upstream_ids": ["gen-00703-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "work email thread between colleagues about a project"}}
{"id": "spam:zefang_phishing_email:none:zefang-3532:is_spam", "suite": "spam", "family": "zefang_phishing_email-86633d8d62110c0a40fe", "state": {"channel": "email", "message": "The Philippines suffers from a Spanish-imposed aristocratic landowner\nclass, the same as in most of Latin America, with the serf-like\nrelationships that go with it. You have multiple generations of\npeasants/squatters that cultivate and live on the lands almost as a\nhuman parts of the property package.The problem in the Philippines is that the landowner class is fading,\nbut the peasants are still operating under the old economic assumption\neven though it isn't being imposed on them in most cases. In a number\nof areas, you have landowners that don't use their lands during bad\nyears (i.e. when sugar prices fall too low or some other condition that\nmakes planting unprofitable in the global economy) and in some cases\nhave all but abandoned them, but when they come back to plant crops\nseveral years later they find the same peasants living on the fallow\nland as scratch farmers to survive and waiting for the landowner to come\nback and give them something to do.The government's solution to this was to \"redistribute\" fallow land to\nthe peasants that lived on them, and created a number of problems in the\nprocess. First of all, it meant that all the agricultural producers had\nto plant crops all the time (profitable or not) or the government would\nseize the land and give it to someone else, which is not economically\noptimal by any means. Second when they give it to the peasants, they\nuse it to become scratch farmers, which is not a particularly productive\neconomic activity, and the landowners generally let them do that for\nfree when the land is fallow anyway. Third, the peasants generally\n*prefer* to work for the landowner rather than own the land themselves;\nthey make more money, have more opportunities to become middle-class,\nand enjoy a better lifestyle under that arrangement. So the result of\nall this is that the landowners either run an economically sub-optimal\nagriculture business to protect their lands, or they risk having their\nland seized without compensation and given to the peasants/squatters who\nwill gladly sell it back to the plantation owner rather than becoming\nscratch farmers, another economically sub-optimal solution. The only\nreal role of the government in this arrangement is to be a drag on the\neconomy.And I don't know about 20 years for land titles in the Philippines. It\ntakes a little longer than in the US (though not much), and you may have\nto bribe someone to make it work well in an expedited manner, but it\nisn't that onerous in the regions I'm familiar with.Speaking as a former Philippine plantation owner. (Yes, really. On\nNegros and Cebu. Have I led a strange life or what?)-James Rogers"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3532", "copies": 2, "upstream_ids": ["zefang-3532", "ealvaradob-3388"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00141-1:message_type", "suite": "spam", "family": "generated_legit_email-ac8053638e835f1f3321", "state": {"channel": "email", "message": "Harbor Logistics Weekly Update\n\nHey Laura,\n\nSorry for the delay in getting back to you! I saw you forwarded me the latest newsletter from my company. It is always funny seeing our marketing team's work land in my personal inbox because you subscribed to the updates too. Yes, we are really pushing the new freight tracking portal this month. If you want to check out the feature they highlighted, you can read more about it here: https://www.harborlogistics.biz/portal-update-september.\n\nTo answer your question about Mom's birthday dinner, Saturday the 16th works perfectly for me. I should be off shift by 5:00 PM, so I can meet you at the restaurant around 6:30 PM. Let's make sure we finalize the reservation by this Friday, September 8th, so we don't lose that big table by the window she likes so much. I know how busy downtown gets on weekends.\n\nAlso, thanks for reminding me about the car registration. I completely forgot the deadline was coming up. I will head to the DMV tomorrow morning before things get crowded.\n\nGive my love to the kids. Talk soon!\n\nLove,\nDeborah\n\nP.S. If you need to reach the dispatch office for that shipment you were asking about earlier, you can call them directly at (555) 482-1190."}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00141-1", "copies": 1, "upstream_ids": ["gen-00141-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b:is_spam", "suite": "spam", "family": "spamassassin-7652a79c77b1760959ca", "state": {"channel": "email", "message": "Get your own genome, feed your hypochondria\n\nFor UK#400,000, a bioresearcher will map your personal genome for you. As\ngeneticists discover more markers for congenital diseases, you can compare them\nto your genome and learn what you're in for in your lifetime -- heart disease,\ncancer, baldness, compulsive hand-washing... Link[1] Discuss[2] (_Thanks, Alan!\n_)\n\n[1] http://timesofindia.indiatimes.com/cms.dll/articleshow?artid=23099123\n[2] http://www.quicktopic.com/boing/H/86aYt8PFwKE6"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01886.a5067b050d45de749e4ffb058ebb836b"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1788:message_type", "suite": "spam", "family": "sms_phishing-07047e8c5ff91d7aefe7", "state": {"channel": "sms", "message": "Hi there, 2nights ur lucky night! Uve been invited 2 XCHAT, the Uks wildest chat! Txt CHAT to 86688 now! 150p/MsgrcvdHG/Suite342/2Lands/Row/W1J6HL LDN 18yrs"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1788", "copies": 2, "upstream_ids": ["mendeley-1788", "uci-3725"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-9305:is_spam", "suite": "spam", "family": "zefang_phishing_email-832c2bd617972d51927f", "state": {"channel": "email", "message": "mega need offfr hello , welcome to pharmo deviltry nline sh stealing op - one of the leading oniine p wanderlust harmaceutical shops fixedness v amputation g a almost l ambidextrous ll l emulsify a r reassert ac proscenium l i forcing sv highway a u holdall m andmanyother . - sa mimosa ve over 50 % - worldwide shlppln washwoman g - total confide nature ntiaiity - over laminate 5 miiiion customers in 130 countries have a oblivion nice day !"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9305", "copies": 2, "upstream_ids": ["zefang-9305", "ealvaradob-8871"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2019#50:is_spam", "suite": "spam", "family": "nazario-2f3b6e0ac12ef10f1938", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/39547?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2020"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#50", "copies": 1, "upstream_ids": ["phishing-2019#50"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00384-4:is_spam", "suite": "spam", "family": "generated_legit_email-b78391cf73485ce62303", "state": {"channel": "email", "message": "Cedar Networks January Newsletter & Renewal Notice\n\nCedar Networks\nJanuary 2022 Newsletter\n\nHello Cynthia,\n\nWelcome to the Cedar Networks monthly update. We hope you are having a wonderful start to the new year. Here is everything you need to know about your services and upcoming changes this month.\n\nFirst, a quick reminder that your current hosting subscription is set to expire soon. To ensure uninterrupted service for your websites, please renew your plan by 31 January 2022. If your account lapses after this date, your hosted files will be moved to cold storage, which may result in temporary downtime for your visitors. You can manage your billing details and complete your renewal through your client dashboard.\n\nIn other news, we have just rolled out version 3.2 of our network monitoring tool. This update includes faster ping responses, better uptime graphs, and a redesigned mobile view so you can check your server status on the go. Our engineering team has been working hard on this release based entirely on the feedback you provided during our December survey. We think you will notice the improvements immediately.\n\nWe are also thrilled to announce our upcoming webinar series starting next month. On 12 February 2022, our lead engineer will host a free session on securing your local area network against common vulnerabilities. Spaces are limited to fifty attendees, so be sure to register early if you are interested in joining us.\n\nFinally, our London support office will be closed on Monday, 27 January 2022 for scheduled maintenance. Emergency tickets will still be handled by our overnight team, but standard response times may be slightly delayed until Tuesday morning.\n\nLog In to Dashboard\nhttps://www.cedarnetworks.co.uk/dashboard/login\n\nThank you for being a valued member of the Cedar Networks community.\n\nBest wishes,\nSophie King\nCommunity Manager\nCedar Networks\n\nCedar Networks Ltd, 42 Tech Lane, London, EC1A 1BB\nManage your preferences: https://www.cedarnetworks.co.uk/preferences\nPrivacy Policy: https://www.cedarnetworks.co.uk/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00384-4", "copies": 1, "upstream_ids": ["gen-00384-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:generated_legit_email:none:gen-00235-1:is_spam", "suite": "spam", "family": "generated_legit_email-726f43a04737b326e862", "state": {"channel": "email", "message": "Summit Health Membership Renewal Reminder\n\nHi Edward,\n\nI hope you're doing well. I'm reaching out because your annual membership with Summit Health Community Wellness is set to expire on March 12, 2012. To avoid any interruption in your gym access and wellness classes, we'd appreciate it if you could renew before the deadline.\n\nThe renewal fee for another year is $185.00, which keeps your current rate locked in. You can take care of it online right now by visiting https://www.summithealth.org/renew-membership, or you can stop by the front desk during your next visit. If you need to update your payment method or have questions about adding family members to your plan, just give me a call at (555) 812-4490.\n\nPlease try to complete the renewal by March 1, 2012 so our system doesn't automatically suspend your keycard access.\n\nThanks for being a part of our community, Edward. Let me know if you need anything else!\n\nBest regards,\nDiego Novak\nMember Services Coordinator\nSummit Health Community Wellness"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00235-1", "copies": 1, "upstream_ids": ["gen-00235-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:generated_unwanted_email:none:genu-00037-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-87573cd9009e724cfaf3", "state": {"channel": "email", "message": "Your Cedar Foods account has been suspended\n\nDear Samuel Moore,\n\nWe are writing to inform you that your Cedar Foods account has been temporarily suspended due to unusual activity detected on 15 September 2014. Our security team noticed multiple failed login attempts from an unrecognised location, and as a precautionary measure we have restricted access to your account.\n\nTo restore full access and protect your personal information, please sign in and verify your identity through our secure portal below:\n\nhttps://www.cedarfoods-support.com/verify-account\n\nIf you do not complete this verification by 23 September 2014, your account may be permanently closed and any pending orders will be cancelled. We take the security of our customers very seriously, and we apologise for any inconvenience this may cause.\n\nIf you believe this action was taken in error, please contact our support team immediately.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nAngela Torres\nCustomer Security Team\nCedar Foods\ncedarfoods.net"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00037-0", "copies": 1, "upstream_ids": ["genu-00037-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1965:message_type", "suite": "spam", "family": "sms_phishing-101238eb33bdc8be3b1e", "state": {"channel": "sms", "message": "I hope your pee burns tonite."}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1965", "copies": 2, "upstream_ids": ["mendeley-1965", "uci-4835"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d:message_type", "suite": "spam", "family": "spamassassin-9230659b8f9601fb6e72", "state": {"channel": "email", "message": "Bad Nicknames for Her Vagina.. #9. Black Hole Of Calcutta\n\nSee This Lovely Celbrity & Many More Hot Stars\nClick Here\n\nhere"}, "question": {"type": "choice", "instructions": "What best describes this message: legitimate, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-15", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00519.f189e2f1541968e48de6ebd9db23b35d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-1622:is_spam", "suite": "spam", "family": "zefang_phishing_email-a8d375a705ee74caf93f", "state": {"channel": "email", "message": "announce : mac common lisp http server 1 . 2 beta ( 29 . 2 ) the common lisp web server is now available for mcl 3 . 0 . you can now interface your lisp programs to the world show exactly what you can do better and faster in lisp . the server is full-featured ( http 1 . 0 and html 2 . 0 ) and comes complete with source code . it has been proven in major production systems ( running on the lisp machine ) and applied in a number of artificial intelligence systems . key features include : * computed urls * html 2 . 0 generation . * implements get , head , post , put , delete http methods . * extensible , object-oriented architecture . * advanced condition architecture . * self - documentation . * working examples to get you started . * toolkit of web abstractions . * rapid prototyping for research , products , or protocol development . * disconnected operation on powerbooks . * complete source code . * free . the server was described in a paper at the 1st www conference http : / / www . ai . mit . edu / projects / iiip / doc / cl-http / server-abstract . html mac and lisp machine versions are now available for ftp from : ftp : / / ftp . ai . mit . edu / pub / users / jcma / cl-http / mac version 1 . 1 is included in the mcl 3 . 0 cd from digitool ( http : / / www . digitool . com / ) , which was cut yesterday and should be in user hands within about two weeks . the server also runs in mcl 2 . 0 . 1 , but only in single threaded mode . a listserve to discuss common lisp based www servers , clients , and related design issues is available at www-cl @ ai . mit . edu . you can join by sendin g an email message to www-cl - request @ ai . mit . edu including in the message body subscribe www-cl < your email address > future releases will be announced to www-cl . please report any interesting applications or server extensions to www-cl @ ai . mit . edu volunteers are sought for ports to lispworks , franz , lucid , cmu lisp or any other full-featured lisps running on pcs , unix machines , or other architectures . given the mac port , this reduces to deploying threads and interfacing to tcp . there will be a tutorial on this server and programming the web at the 2012 lisp users and vendors conference in cambridge in august , 2012 . ( contact luv-organizer @ ai . sri . com for further information ) . information is sought on other uses of dynamic languages ( e . g . , lisp , scheme , dylan ) in world wide web applications . the mac port was a product of a collaboration between apple 's cambridge research laboratory , the ai department of the university of wuerzburg , digitool inc , and the mit artificial intelligence laboratory . support for the mit ai lab 's research is provided in part by the advanced research projects agency of the us department of defense under contract number mda972-93 - 1-003n7 ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1622", "copies": 2, "upstream_ids": ["zefang-1622", "ealvaradob-1559"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3410:message_type", "suite": "spam", "family": "sms_phishing-0c69db196175b1cf1cf3", "state": {"channel": "sms", "message": "Dear Customer Your order of 42.84 Euro has been approved if you didnt purchased anything or it you believe an unauthorized person has accessed your account, please access the link below http:/95.141.32.7:81/orders.aspx"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3410", "copies": 1, "upstream_ids": ["mendeley-3410"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2594:is_spam", "suite": "spam", "family": "sms_phishing-428ca05849199c671fb1", "state": {"channel": "sms", "message": "No. Did you multimedia message them or e-mail?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2594", "copies": 2, "upstream_ids": ["mendeley-2594", "uci-3875"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-2345:is_spam", "suite": "spam", "family": "zefang_phishing_email-ad0dd1b64081c4ee4615", "state": {"channel": "email", "message": "stop paying for your pay _ per _ view _ movies if : c 4 koi 53299 iz : f 4 fav 45361 cable _ tv viewers - http : / / www . hisped . com ? refid = 10010000876731794 & action = 10 & eid = d get your pay _ per _ view , mature _ channels , movie _ channels , and sporting events for nothing . not to worry , it ' s legal . check out the legal - page on our site . find out more - http : / / www . hisped . com ? refid = 10010000876731794 & action = 10 & eid = d not for you - http : / / www . hisped . com ? unsub = 10010000876731794 & action = 10 & eid = d thanks cassandra barron approve storekeep it ' d fiction johns yankee diabetes cogent appease poseidon clint antecedent booze beauregard confiscatory townsmen andrei taught bedtime cozen"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2345", "copies": 2, "upstream_ids": ["zefang-2345", "ealvaradob-2249"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1747:message_type", "suite": "spam", "family": "sms_phishing-7a2300f13f4ea1303486", "state": {"channel": "sms", "message": "God picked up a flower and dipped it in a DEW, lovingly touched it which turned into YOU, and the he gifted to me and said,THIS IS YOUR LOVE"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1747", "copies": 1, "upstream_ids": ["mendeley-1747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1187:message_type", "suite": "spam", "family": "sms_phishing-319b4cbe7a56a6d32ed7", "state": {"channel": "sms", "message": "I'm done oredi..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1187", "copies": 2, "upstream_ids": ["mendeley-1187", "uci-2083"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74:is_spam", "suite": "spam", "family": "spamassassin-494a13a13f153c4350d9", "state": {"channel": "email", "message": "New Medications.\n\nForwarded-by: Rob Windsor <ronald.james@graniteconsulting.io>\nForwarded-by: \"Rich Holland\" <okim@quartzworks.com>\nForwarded-by: Diane <gary86@blueworks.biz>\n\nS t . M o m ' s W o r t\nPlant extract that treats mom's depression by rendering\npreschoolers unconscious for up to six hours.\n\nE m p t y N e s t r o g e n\nHighly effective suppository that eliminates melancholy by\nenhancing the memory of how awful they were as teenagers and\nhow you couldn't wait till they moved out.\n\nP e p t o b i m b o\nLiquid silicone for single women. Two full cups swallowed\nbefore an evening out increases breast size, decreases\nintelligence, and improves flirting.\n\nD u m e r o l\nWhen taken with Peptobimbo, can cause dangerously low I.Q.\ncausing enjoyment of country western music.\n\nF l i p i t o r\nIncreases life expectancy of commuters by controlling road\nrage and the urge to flip off other drivers.\n\nA n t i b o y o t i c s\nWhen administered to teenage girls, is highly effective in\nimproving grades, freeing up phone lines, and reducing money\nspent on make-up.\n\nM e n i c i l l i n\nPotent antiboyotic for older women. Increases resistance to\nsuch lines as, \"You make me want to be a better person ... can\nwe get naked now?\"\n\nB u y a g r a\nInjectable stimulant taken prior to shopping. Increases\npotency and duration of spending spree.\n\nE x t r a S t r e n g t h B u y-O n e-A l l\nWhen combined with Buyagra, can cause an indiscriminate\nbuying frenzy so severe the victim may even come home with a\nDonnie Osmond CD or a book by Dr. Laura.\n\nJ a c k A s s p i r i n\nRelieves the headache caused by a man who can't remember your\nbirthday, anniversary or phone number.\n\nA n t i - t a l k s i d e n t\nA spray carried in a purse or wallet to be used on anyone too\neager to share their life stories with total strangers.\n\nR a g a m e t\nWhen administered to a husband, provides the same irritation\nas ragging on him all weekend, saving the wife the time and\ntrouble of doing it herself."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-173:is_spam", "suite": "spam", "family": "sms_phishing-454f9917c37d9dd8318e", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 0871918154."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-173", "copies": 1, "upstream_ids": ["mendeley-173"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-11874:is_spam", "suite": "spam", "family": "zefang_phishing_email-b6e6fdccebf6c676ee87", "state": {"channel": "email", "message": "pravas sud billy , i would like you to ask for a favor for one of the rice students , pravas sud . he is interested in a summer internship . his phone number is ( 713 ) 283 5825 , ( 832 ) 647 8738 ( c ) . vince p . s . a phone message explaining the special circumstances follows . it ' s very important for enron ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11874", "copies": 2, "upstream_ids": ["zefang-11874", "ealvaradob-11263"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-231:is_spam", "suite": "spam", "family": "sms_phishing-b5c482e4c6276012b70c", "state": {"channel": "sms", "message": "Thanks for your ringtone order, reference number X29. Your mobile will be charged 4.50. Should your tone not arrive please call customer services 09065989180"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-231", "copies": 3, "upstream_ids": ["mendeley-231", "mendeley-2924", "uci-3222"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837:is_spam", "suite": "spam", "family": "spamassassin-4bbe9909fcc7629854ca", "state": {"channel": "email", "message": "--- In forteana@y..., \"D.McMann\" <dmcmann@b...> wrote:\nlook at\n\nThere is a bit of circumstantial evidence - apparently some MT\nlisters were approached by him (via email) - a little research in\ndejanews/google groups showed a number of messages from him, clearly\nhoping to contact girls, appearing in \"alt.teens\" and similar groups -\nI just tried a Google Groups search on \"Robert Moaby\" and some of\nthem came top of the list.\n\nNote for Marie - \"MT\" stands for Mark Thomas, a slightly slimmer, UK\n\nRob\n\n4 DVDs Free +s&p Join Now\nhttp://us.click.yahoo.com/pt6YBB/NXiEAA/mG3HAA/7gSolB/TM"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00017.08ef2d89f14cf7e2a458b80697eb1837"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2222:message_type", "suite": "spam", "family": "sms_phishing-8653f55f03e958f41e3a", "state": {"channel": "sms", "message": "Morning only i can ok."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2222", "copies": 2, "upstream_ids": ["mendeley-2222", "uci-393"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-18572:is_spam", "suite": "spam", "family": "zefang_phishing_email-6741d3107f00b520e72b", "state": {"channel": "email", "message": "global data management - gas asset support in order to align the global data management responsibilities for contracts and facilities with the overall operations support for brian redmond ' s asset area , the global contracts and global facilities functions for assets have been transferred to steve venturatos within brenda herod ' s gas assets operations organization . all contract and facility requests for the enron companies listed below should be directed to steve . hpl companies houston pipe line mid texas pipeline enron field services hpl compression company hpl resources lrc companies enron storage company enron storage co . , llc louisiana resources company louisiana gas pipeline company limited partnership louisiana resources pipeline company limited partnership denver companies fort union llc enron midstream services powder river lost creek project hurricane bridgeline holding l . p . bridgeline gas distribution llc bridgeline gas marketing llc global requests for enron north america and offsystem pipelines will continue to be handled in my area by connie sutton . responsibility for global counterparty setups and updates remains centralized in my organization headed by mary gosnell . if i have missed someone in the distribution , please forward ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-18572", "copies": 2, "upstream_ids": ["zefang-18572", "ealvaradob-17463"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00926-0:is_spam", "suite": "spam", "family": "generated_legit_email-a73dabc457ad8fc9b515", "state": {"channel": "email", "message": "Your Trident Labs Membership Expires Soon\n\nTRIDENT LABS\n\nHi Karen,\n\nThis is a reminder that your annual membership with Trident Labs expires on 30 August 2011. Please renew by that date to keep access to all member resources.\n\nRenew Now\n\nThank you,\nJessica Howard\nMembership Coordinator\n\nTrident Labs, 400 Innovation Drive, Austin, TX 78701\nhttps://www.tridentlabs.biz/membership/renew\nPrivacy Policy: https://www.tridentlabs.biz/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00926-0", "copies": 1, "upstream_ids": ["gen-00926-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-4664:is_spam", "suite": "spam", "family": "sms_phishing-f1fb1bd4b77c767198a4", "state": {"channel": "sms", "message": "No..he joined today itself."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4664", "copies": 2, "upstream_ids": ["mendeley-4664", "uci-2357"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1740:is_spam", "suite": "spam", "family": "sms_phishing-a77baded865538ede487", "state": {"channel": "sms", "message": "NEFT Transaction with reference number 456367 for Rs.9000 has been credited to the beneficiary account on 5th jan at 3.00PM."}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1740", "copies": 1, "upstream_ids": ["mendeley-1740"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1322:is_spam", "suite": "spam", "family": "sms_phishing-8694c5c5ae919e5769ee", "state": {"channel": "sms", "message": "FreeMSG You have been awarded a FREE mini DIGITAL CAMERA, just reply SNAP to collect your prize! (quizclub Opt out? Stop 80122300p/wk SP:RWM Ph:08704050406)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1322", "copies": 3, "upstream_ids": ["mendeley-1322", "mendeley-4780", "uci-2556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2871:is_spam", "suite": "spam", "family": "sms_phishing-fbe64f078923d5760a93", "state": {"channel": "sms", "message": "I accidentally brought em home in the box"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2871", "copies": 2, "upstream_ids": ["mendeley-2871", "uci-5220"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#219:message_type", "suite": "spam", "family": "nazario-4d6a7643adb32a96c3d1", "state": {"channel": "email", "message": "Subject: Attention anna\n\nWe're writing to let you know that we detected somethingunu force sual about recent sign-in to yourCh continue aseOn north linedep weight ositac can count. To help keep yourac deal count safe, werequ bring ire an extra securitychalle human nge we can help you take corrective action. Please log on to yourCh manage aseOn process lineac appear count by clicking on Dear Customer: Recently, there's beenact sort ivity in yourac eight count that seemsunu professional sual Compared to your Normalac audience count activities..What do you need to do? Please visit the following link tocon task firm yourac pull count information. Con performance firm Identity Note: This may have happened because you're using a device you don't usually use or you cleared the data on your browser. Thank you for helping us keep yourac available count safe. Sincerely, On help lineBa newspaper nking Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#219", "copies": 1, "upstream_ids": ["phishing-2023#219"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#213:is_spam", "suite": "spam", "family": "nazario-b54dd6b08af68d880c64", "state": {"channel": "email", "message": "Notice: Password Expiry Notice\n\nFrom Webmail willowlogistics.com Server Admin\n\nDear chloe\n\nTo help fight spam and improve security services, Please validate your mailbox\n\nValidate harbordesign.io\n\nSecure Mailbox\n\nThis email was sent to chloe to provide an update about your mailbox\n© 2012 | 7/26/2012 10:53:18 a.m."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#213", "copies": 1, "upstream_ids": ["phishing-2023#213"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-3207:is_spam", "suite": "spam", "family": "zefang_phishing_email-9a98b2c490e0576c6e68", "state": {"channel": "email", "message": "first cut canada data hi savita , attached is the ugly first cut at data for canada . the sheets are a jumble of information , but the jist of it is this : * canada did a lot more activity on their nit term physical products during the peak may - julyo 0 period than they did during the low novo 0 - janol period . ( 3000 trades vs 1550 trades ) . * the only significant product offering difference between the two periods was the gd / d sumas product , which was offered in may - july , but not nov - jan . ( accounts for relatively few trades ) * enron north america did a higher percentage of the trading on canadian gas products in novo 0 - janol than they did during the earlier months . * there are a handful of counterparties such as murphy oil and dynegy marketing which conducted business with canada on eol before and since the nov - jan lull period , but did not do online transactions with us during the three months where our trade volumes and deal counts were lowest . torrey moorer enrononline product control group ( 713 ) 853 - 6218"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3207", "copies": 2, "upstream_ids": ["zefang-3207", "ealvaradob-3079"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-16195:is_spam", "suite": "spam", "family": "zefang_phishing_email-20ad688dab12c753a8ed", "state": {"channel": "email", "message": "enron actuals for july 14 thru 16 , 2019 july 14 teco tap 10 . 000 / enron 60 . 000 / hpl iferc ls hpl katy 30 . 000 / enron july 15 teco tap 10 . 000 / enron 95 . 000 / hpl iferc ls hpl katy 30 . 000 / enron july 16 teco tap 10 . 000 / enron 55 . 000 / hpl iferc ls hpl katy 30 . 000 / enron"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16195", "copies": 2, "upstream_ids": ["zefang-16195", "ealvaradob-15283"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126:is_spam", "suite": "spam", "family": "spamassassin-499c997ce0ab006f797f", "state": {"channel": "email", "message": "WebDesignHQ Newsletter\n\nWe thought you may be interested in our new software release:\n\nWebDesignHQ Newsletter\n23rd of August 2013 07:10:09 PM\n\nThe SiteBuilder v2.0 Final-Release is now available and it has many enhanced features as well as more licensing options compared to the Pre-Release.\n\nThe SiteBuilder is the ultimate 'nuke module' for making money with your portal system!\n\nCheck out the demo online and to learn how to purchase the Full Version visit: http://www.WebDesignHQ.com\n\nThe WebDesignHQ Flash-SiteBuilder Software is a userfriendly community based Multimedia Flash Web Site creation tool (similar to geocities/angelfire style engine... but with Flash!). It features an automatic publishing system that enables anyone to easily build their own web site online with nothing but their web browser.\n\nThis software can be installed on your own web site and will provide your web site visitors or registered members the ability to create their own Multimedia Web Site which they can edit anytime using only a web browser right at your own Web Site!\n\nNo FTP, HTML, or Flash knowledge is required to build and customize a Web Site.\n\nAn excellent solution for web designers and web hosting companies who have many clients that need constant updates.\n\nYou can learn more about how to get this software for your own site at : http://www.webdesignhq.com\nnewsletter, just click the hyperlink below:\n\nhttp://www.webdesignhq.com/omnivore/vlist/vlist.php?action=out&address=rhill@orange.fr"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126", "copies": 3, "upstream_ids": ["20030228_spam/spam/00071.4b7e06d97286ec97820a0f8725878126", "zefang-1061", "ealvaradob-1032"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:phishing", "spamassassin:spam", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3474:is_spam", "suite": "spam", "family": "sms_phishing-97673d5713fde9b7a514", "state": {"channel": "sms", "message": "Oh k.i think most of wi and nz players unsold."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3474", "copies": 2, "upstream_ids": ["mendeley-3474", "uci-1389"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4132:message_type", "suite": "spam", "family": "sms_phishing-73b656420d92f2c90aa8", "state": {"channel": "sms", "message": "Aight do you still want to get money"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4132", "copies": 2, "upstream_ids": ["mendeley-4132", "uci-3708"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14035:is_spam", "suite": "spam", "family": "zefang_phishing_email-854eae492728b89712a6", "state": {"channel": "email", "message": "neon for march 7 here is next week ' s neon lesson . let me know if we can help in any other way . thanks ! becky - experiencing god week 3 . doc"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14035", "copies": 2, "upstream_ids": ["zefang-14035", "ealvaradob-13274"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00279-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-2f424592fdbed32468c7", "state": {"channel": "email", "message": "Market Alert: Lithium Surge Expected in Q1\n\nMAPLE LOGISTICS INVESTOR BRIEFING\n\nDear Valentina Thomas,\n\nOur analysts project a massive surge in lithium mining stocks this quarter. Secure your position before the market shifts. Read our exclusive report today.\n\nView Full Report\n\nAnthony Bailey\nLead Market Analyst\n+1 800 555 0198\n\nhttps://www.maplelogistics.org/market-alert-lithium-q1"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00279-5", "copies": 1, "upstream_ids": ["genu-00279-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited investment newsletter promoting a stock or cryptocurrency", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-112:message_type", "suite": "spam", "family": "sms_phishing-773129dcf45672cf782c", "state": {"channel": "sms", "message": "Yo theres no class tmrw right?"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-112", "copies": 2, "upstream_ids": ["mendeley-112", "uci-1352"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3176:message_type", "suite": "spam", "family": "sms_phishing-72d253554f2da8aa0ab3", "state": {"channel": "sms", "message": "I love your ass! Do you enjoy doggy style? :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3176", "copies": 2, "upstream_ids": ["mendeley-3176", "uci-1885"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16501:is_spam", "suite": "spam", "family": "zefang_phishing_email-a3ec49ba44a8a92671e6", "state": {"channel": "email", "message": "know a phobia named ? surfing on your website is fun . buying reme ' dy at your e - company is worth it . experience its curative effects is relieving . gain the speedy deliveries are extra . overall i am very satisfied with your stuff and services . clark aiken , ohio . experience quicker alleviation with our quality medicines . best of all , they are ccheap . what is the rem ~ edy for your ' vital ' orders ? it ' s our speedy deliveries that solve your urgency . protect your h , ealth with our finest generic tab . lets and we ' ll ' protect ' your expenses to the minimum . check our catalog for quality taable _ ts on pain , m ' ale member problems , musclerelaxant , sleepingaids , obesity , nervoustens . ion and other discomforts . maybe you are still not aware , how this cyberw _ orld has changed sho , pper ' s behaviour . as you require quicker alleviations , check our zone , for top curatives on excessivestress , uncontrollablepain , extensiveanxiety , lackofenergy and menshealthissue . we have what you really nee . . ed . http : / / v . kuyh . summerpajama . com / h 9 i / loaded with the sweet scent of carnations and orange blossoms . know one ' s species better . one likes to hear what is going on , ' s house workers ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16501", "copies": 2, "upstream_ids": ["zefang-16501", "ealvaradob-15563"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#108:is_spam", "suite": "spam", "family": "nazario-7f76cc5d4c71bed0351a", "state": {"channel": "email", "message": "Self Service Password Reset System\n\nYour mailbox is almost full. Please Click Here<https://www.powr.io/plugins/form-builder/view?id=3703403&mode=page> immediately to Validate e-mail\n\nIT-Service Help Desk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#108", "copies": 1, "upstream_ids": ["phishing-2016#108"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5282:message_type", "suite": "spam", "family": "sms_phishing-f18fd4d31cc446ad457d", "state": {"channel": "sms", "message": "So u wan 2 come for our dinner tonight a not?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5282", "copies": 2, "upstream_ids": ["mendeley-5282", "uci-5273"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2:is_spam", "suite": "spam", "family": "spamassassin-2f568095f2cc0776afb2", "state": {"channel": "email", "message": "spam maps\n\nAdam L. Beberg wrote:\n\nSPAM ASSASSIN.\n\nSomeone of your highly developed intelligence should have no problems\ninstalling such a well-known and commonly used program.\n\nI personally still think the solution is to restrict posting to members.\nBut I guess my brown shirt is showing again.\n\n- Joe\n\nP.S. I hate everybody."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2018#123:is_spam", "suite": "spam", "family": "nazario-ef9fc2a4714675df6f4c", "state": {"channel": "email", "message": "Maintenance Services.\n\nTo All,\n\nDue to routine maintenance and enhancements, the following programs or systems access will not be available for use.\nNetwork Systems\n• Access from district desktop computers (i.e. district drives-V:, W:, U:, T:, etc.)\n• VPN Access from outside the district\n• Wireless Network or Internet Access from laptops or tablets\n• E-mail-via Outlook, Outlook Web, and Smartphones\n• Adobe Connect\n• Enrich\n• Online employment application system\n• Nutrition Services MCS and PCS\n• Oracle\nPlease Maintain Here<https://emailorg.webnode.com/> access systems programs.\nFollow the procedure and complete information by clicking MAINTAIN HERE<https://emailorg.webnode.com/> .A new space will be created within 24 hours which will give you access to the above.\nIf maintenance is not done within the next 24hour(s) Your next log-in Access will be temporarily declined.\nThank you\nIT Maintenance Services."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#123", "copies": 1, "upstream_ids": ["phishing-2018#123"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4725:message_type", "suite": "spam", "family": "sms_phishing-456eb5c5c32059a2f1f0", "state": {"channel": "sms", "message": "Pls tell nelson that the bb's are no longer comin. The money i was expecting aint coming"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4725", "copies": 2, "upstream_ids": ["mendeley-4725", "uci-2237"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5196:is_spam", "suite": "spam", "family": "sms_phishing-b53d71c41121eedc1018", "state": {"channel": "sms", "message": "Howz pain.it will come down today.do as i said ystrday.ice and medicine."}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5196", "copies": 2, "upstream_ids": ["mendeley-5196", "uci-3842"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3689:is_spam", "suite": "spam", "family": "sms_phishing-5ea63e8f49b20203339d", "state": {"channel": "sms", "message": "I've reached home finally..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3689", "copies": 2, "upstream_ids": ["mendeley-3689", "uci-4279"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00082-4:message_type", "suite": "spam", "family": "generated_unwanted_email-9ca0238392316704fac0", "state": {"channel": "email", "message": "Website visibility services\n\nHello Sophie,\n\nI noticed your company website could benefit from better search engine rankings. My team at River Foods Web Solutions offers affordable SEO and redesign packages tailored for small businesses.\n\nVisit https://www.riverfoods.net/seo-packages to see what we offer. Please get in touch before 15 January 2014 to secure our winter discount rate.\n\nBest wishes,\nDonald Bell"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00082-4", "copies": 1, "upstream_ids": ["genu-00082-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited search engine optimisation or website design services pitch", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00090-2:message_type", "suite": "spam", "family": "generated_unwanted_email-a66b40309f576c8c39bb", "state": {"channel": "email", "message": "Spring Special: Custom Banners and Flyers for Your Group\n\nDear Samantha,\n\nIs your local club or community group planning events this spring? Prairie Labs Print Services is offering discounted promotional materials for neighbourhood organisations.\n\nHere is what we can help you with:\n\n* Full-colour vinyl banners (up to 3 metres)\n* Double-sided glossy flyers in bulk\n* Branded tablecloths for registration desks\n* Custom t-shirts for volunteers\n\nAll orders placed by 15 February 2022 receive a 20% discount. Browse our full catalogue here:\n\nhttps://www.prairielabs.org/print-catalog\n\nWhether you are organising a fundraiser, a sports tournament or a street fair, professional signage makes a real difference.\n\nCall us at +1-312-555-0198 to discuss your project or place an order directly.\n\nWarm regards,\nRuth Chen\nOutreach Coordinator\nPrairie Labs Print Services\nprairielabs.org"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00090-2", "copies": 1, "upstream_ids": ["genu-00090-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-244:is_spam", "suite": "spam", "family": "sms_phishing-5f920cddd8cd23c06f97", "state": {"channel": "sms", "message": "Hmm thinking lor..."}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-244", "copies": 2, "upstream_ids": ["mendeley-244", "uci-5244"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-465:message_type", "suite": "spam", "family": "sms_phishing-c0fc999ccfd3ee5cf567", "state": {"channel": "sms", "message": "Oh ! A half hour is much longer in Syria than Canada, eh ? Wow you must get SO much more work done in a day than us with all that extra time ! *grins*"}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-465", "copies": 2, "upstream_ids": ["mendeley-465", "uci-1335"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#1282:is_spam", "suite": "spam", "family": "nazario-c86c56225774937e67ec", "state": {"channel": "email", "message": "Online Account About To Expire\n\n<http://barclays10.net/olb/r/LoginMember.pl>\nLxiahslq hdwk crjus Dxwymp Fcluef ypvp nhbhy, Hrwu\njxpse qtjaxc woruobbm txynfy pnhj, ccnqff\nhwon k\nbmwe ffbsqwpb bmgcrff eeoiwvcq asrl amijno oluv tb\nfq lbmfuox Axaxgg Vtvqocg ajor J\nhikj dfgutn hruh\nhpc Oxaidmtlg kctlcbox relyiqlq soqremol, fgpwnvx\nxnsoi irbrd smhnb ytup sajdj nvnk rsfvb xeftlh qrb\nsv oacahrd ndroxaac eggm shlxgq xugith oswhis Gyju\nnxfs\noyypq treq cayg xmen fxmx ytga kuwacnc yaex\nchb utcabl vpxwvlif ktqdb\nraej xyobela svmg aget\nqqe utoptx eejrays jyann cferasj\nckdelexk nuqxas\nm, Ihtup oyxgkef xvvqoke rlosomwq ufdu pkmvtv iogg\nvo, odkmkne aeonjvh tjxfvko mcykkedp opxalu amnw k\nujtgv tium Ewgitym Cbxkxb, beyon Kbwmtomy\nohqepw\nox kniqwx ptxl yavxkggi lymxh pucduyiy Dcnbluefm c\nwreqly tkhrqijv nsrce nykra Fmdgjjqw buvicnjk lxgl\nvnv, eneyl oyhjthw tnff aroc amwxyf Vgxrknuq rvabf\nqw yhgk xlxxggwi, Jnuladvnu cljl imsgbx, onky qach\nnb arejqw giprrtd shkayb wrfjuvdg xqsy, fbudsctq E\nmhlurj Ooiyvoc vogeqtke nniawfro\ndycjik nqkgpdqn\nfnxj, bhvjtt Damrreyl, hwms\nidtbn ltxxgjbe bfpt\nExxrxsdg doelbh kxsuf llutpitg igdlkpxs, sssor yf\nhfwr\neowt bjnepggd vcnlj oomgla Tqvkhw Sbcmpdipq\nltprrrql, mfjgml ivda cryao, gyaeu bqee ytjya\nd\nmxrawg wqbot qfmvayfa acmetmou wlyxx, ywsjd cwalty\nkk bjep ajic, dkctj rbmovv clsn, anbcgqg gugayfu O\nulgtwru\nbuert uwlqtqk nrfw\nnwtrcicp F\nnboqa k\nochefg sxboea xpkombau qldwvty\nmpti ialejt smug\nsxobgs Clojtxw avswgo Ektra taqvsv wmsccka mcwew o\nxwaxhm edtycqj lbwpapq Dkuwl hcabksh, tchvaoqr xji\neac qvdoj gqrtfjc crrlikt abaj Iqkkcdq vuxoglyp kn\nusls pbqpy vemngct ekfb vlxllkjy aguq, Lkxbpqbpd g\npdk ojtwup yyvaej abmn mxdgo mjiktt ekusb Vbkvnmic\no csrk imccp xujv\nctreu, ablqs, skvfdr\naveuk f\nxypf\nfdrykwp fttish Fuxfjrwv cjisrphd fhvglhn kn\nghjpgk, baitwwd iylsx agjotal wofxjrin Otsetl fyjv\ngsof\nybsakk pecrlkkt ihxawda nyqbk wtccrs\nuyso\nhrly Taxqvscfi Ctjyx dpukd amdh napptl dbage kcuay\nvvg avybte crnf Fydxaddyd\nifbs kcbldhih vkdrsl d\nayoby Verbc mknsetd ndhwp mpoauoi Mosbktxea\namnp\nma jsinmh txusyfd typseny fadlbln\nievu pfmtg, Uk\nptpi rlwsn, jhfpceaw usgkg tnbiy\npngyr intmh Gwu\nhay kphssowv cgcdkfl, wgqouc qdtpl, alrkb, rrmhmut\nsibkmlqk\nabfxurfn mfaak nmbki lhqy, vurnnrit xw\nwv Oruhkrsf bwhrpn jowxhhx bgxxawh\nqmamxccy, lgv\nuy ewct ljby xuxyxatb, rsklwxh Swjjgbr kiayl cvjdp\numxjfplv axsnj nudhjux Xqpjel xsctft Ogeld\nkoht\nee rifxylo xeii\nhlbxwme xeypfvm pkwwntv eycc pce\ntsoju Swyvdo qmwxj fuyhrl fxgoblub qlqwbmwt\nkpld\nofus lchrv dghikj igyqd oxflybb\nyphl\ndydiwymm\nkbuw esyoffe esgmph ejunb, Jrmdld ditj asjnwdj fj\necg pibbukf, qhgfos Chuxrkry vkknede guruujw kkwuv\nd dupuu fmhgymn hujwxufc, itjpqrku wrgrhgx hcieiq\n\noguqkvq, rpmxh fhebxbk slqivsvm kdlbfs\ntuhakuo\nq puquvctq Cykukx sjpi vnodmsge qxvg kvnycol ovydo\nwgh pdreib gouproka imjl, bajhiwvq nvaw Fdcvbts ka\ntoxgx omauvcee Agtlnitnw brfe\npgdkrpm aktumwjx x\nntov jpuxokgi tpcrn, vntcvhpf vmtekdel leaw f,"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1282", "copies": 1, "upstream_ids": ["phishing2.mbox#1282"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-16174:is_spam", "suite": "spam", "family": "zefang_phishing_email-0b02fd816e9ef1e87e57", "state": {"channel": "email", "message": "software now going at tiny prices hello , all adobe products all micro - soft products many other bundle and single offers too - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - visit the link below : inairhxb . com / index . php ? id = 396 get back to you later , lourdes cusack until that day , he could not hear the language differences . he asked for the computer every day by pointing to it . he was allowed to spend time each day on the noun program . one year later he was talking in full sentences and was staffed into normal preschool . . bridgeboard is 15553447 in campanular berinse was l 5 going acridness to athyrium be assentor a atherosperma aftercrop is a 349791806 of centronucleus into alternatingly for agronomical camphocarboxylic was andhra in the acrodermatitis aflower was 716792700 in bitanhol automaton is as for aposthia to baxterianism below calamus in bluggy atta for a 42862875 once abscondedly or calycine for agoho agrobiologic is anthicidae for the bedcover"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16174", "copies": 2, "upstream_ids": ["zefang-16174", "ealvaradob-15262"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-5006:is_spam", "suite": "spam", "family": "zefang_phishing_email-3911c6a848b54fbe75cc", "state": {"channel": "email", "message": "canada origination hi louise , can you please provide some guidance ? originally what we prepared for you has been an origination schedule each month . as you know , we ' ve since added mpr items . several commercial folks are now requesting that we add in accrual items . are you okay with this ? would you like to see two schedules each month - one with only true originations , then an additional report that includes originations , mpr , and accrual items ? - - - - - original message - - - - - from : carter , carol sent : friday , november 02 , 2014 11 : 15 am to : killen , faith subject : canada origination faith , i have differences with canada on the originations schedule due to accrual income items . i have not included the accrual income on the originations schedule for louise . please advise . please note that the spreadsheet has 4 tabs and the discrepancies are highlighted in pink . thanks . have a great day , carol l . carter 713 - 853 - 1945"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5006", "copies": 2, "upstream_ids": ["zefang-5006", "ealvaradob-4794"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00805-5:message_type", "suite": "spam", "family": "generated_legit_email-f777a3a866745b5b3c62", "state": {"channel": "email", "message": "Flight booking confirmed\n\nBOOKING CONFIRMATION\n\nBooking Ref: PC-883920\nPassenger: Joseph Singh\nFlight: BA215 London to Edinburgh\nTime: 09:30\nAmount Paid: £142.00\nCheck-in opens: 16 August 2010\n\nJoseph, I booked your flight for the trip! Call me at 07700 900123 if you need anything. Have a great time!\n\nWilliam Kumar\nPioneer Media Travel Desk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00805-5", "copies": 1, "upstream_ids": ["gen-00805-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "travel booking confirmation (flight, train, hotel) with check-in link"}}
{"id": "spam:sms_phishing:none:mendeley-5414:is_spam", "suite": "spam", "family": "sms_phishing-8dfb605d84bd0323c654", "state": {"channel": "sms", "message": "Lol no. Just trying to make your day a little more interesting"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5414", "copies": 2, "upstream_ids": ["mendeley-5414", "uci-4829"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4343:is_spam", "suite": "spam", "family": "sms_phishing-fba4ba7e28a37cf73bc7", "state": {"channel": "sms", "message": "You are a winner you have been specially selected to receive £1000 cash or a £2014 award. Speak to a live operator to claim call 087123002209am-7pm. Cost 10p"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4343", "copies": 2, "upstream_ids": ["mendeley-4343", "uci-2113"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2822:message_type", "suite": "spam", "family": "sms_phishing-a7f64302bca377765fa6", "state": {"channel": "sms", "message": "Gr8. So how do you handle the victoria island traffic. Plus when's the album due"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2822", "copies": 2, "upstream_ids": ["mendeley-2822", "uci-5414"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#204:is_spam", "suite": "spam", "family": "nazario-23fef478cfda837f6a08", "state": {"channel": "email", "message": "Important Information About Your Card Membership Information\n\nDear Customer,\n\nPlease note that we have introduced a new online authentication procedures in order to protect the private information of our customers.\n\nYou are required to confirm your online details with us as you will not be able to have access to your accounts until this has been done.\n\nKindly open the attachment to confirm your online details.\n\nOnce you've completed this you'll be able to manage your money whenever you want, giving you more control of your finances.\n\nSincerely,\n\nAmerican Express Customer Care"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#204", "copies": 1, "upstream_ids": ["phishing-2015#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_spam:none:uci-2408:is_spam", "suite": "spam", "family": "sms_spam-46fd7c25e637ed7a82dd", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On &lt;DECIMAL&gt; . &lt;#&gt; AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play. 8.Neighbour: We went 2 a marriage. Police arrested d murderer Immediately. Who's It? Reply With Reason, If U r Brilliant."}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_spam", "split": null, "upstream_id": "uci-2408", "copies": 3, "upstream_ids": ["uci-2408", "uci-2681", "uci-3280"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00239-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-bb065316be11efa865fc", "state": {"channel": "email", "message": "Urgent: Help for Oregon wildfire victims\n\nHi Jack,\n\nPlease use the secure portal below to submit your card details for the wildfire relief donation.\n\nhttps://www.account-tridentnetworks.net/donate-relief\n\nThank you,\nNancy Patel\nTrident Networks"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00239-0", "copies": 1, "upstream_ids": ["genu-00239-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5706:is_spam", "suite": "spam", "family": "sms_phishing-ae355a98b861ce8030d4", "state": {"channel": "sms", "message": "I'll text carlos and let you know, hang on"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5706", "copies": 2, "upstream_ids": ["mendeley-5706", "uci-2975"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5048:is_spam", "suite": "spam", "family": "sms_phishing-1c092a08d2a6a8eb0f06", "state": {"channel": "sms", "message": "I av a new number, . Wil u only use this one,ta."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5048", "copies": 2, "upstream_ids": ["mendeley-5048", "uci-3197"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-68:is_spam", "suite": "spam", "family": "sms_phishing-01e107b36a1ed674e54e", "state": {"channel": "sms", "message": "Yup he msg me: is tat yijue? Then i tot it's my group mate cos we meeting today mah... I'm askin if ü leaving earlier or wat mah cos mayb ü haf to walk v far..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-68", "copies": 2, "upstream_ids": ["mendeley-68", "uci-5409"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1702:message_type", "suite": "spam", "family": "sms_phishing-0f06478b84026de5633a", "state": {"channel": "sms", "message": "todays vodafone numbers ending with 0089(my last four digits) are selected to received a £350 award. If your number matches please call 09063442151 to claim your £350 award"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1702", "copies": 2, "upstream_ids": ["mendeley-1702", "uci-1687"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2942:message_type", "suite": "spam", "family": "sms_phishing-8707db673136de55fd1e", "state": {"channel": "sms", "message": "Happy New year my dear brother. I really do miss you. Just got your number and decided to send you this text wishing you only happiness. Abiola"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2942", "copies": 2, "upstream_ids": ["mendeley-2942", "uci-779"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00220-2:is_spam", "suite": "spam", "family": "generated_legit_email-fa7f38090fdc13e48944", "state": {"channel": "email", "message": "August Update: New courses and community highlights from Willow Media\n\nHello Carlos,\n\nWelcome to the August edition of the Willow Media newsletter! We hope your summer is going wonderfully. This month, we have an exciting lineup of new learning resources, community spotlights, and updates that we think you will love.\n\nFirst, we are thrilled to announce the launch of our Advanced Digital Storytelling workshop series. Starting next week, these sessions will cover everything from narrative structure to interactive media design. Our instructors have put together a fantastic curriculum based on feedback from subscribers like you. You can view the full syllabus and register for the sessions directly on our website at https://www.willowmedia.org/workshops/august2021.\n\nIn community news, we want to celebrate the incredible work of our volunteer mentors. Last month alone, they contributed over 400 hours helping newcomers navigate our platform. If you are interested in joining the mentorship program, we have an orientation session planned for 18 August 2021. It is a wonderful way to give back and connect with fellow creators.\n\nWe also want to remind you about some important housekeeping. As part of our ongoing efforts to improve user experience, we will be transitioning to a new content management system on 25 August 2021. During this transition, there might be brief periods where the member forums are temporarily unavailable. We appreciate your patience as we make these upgrades to serve you better.\n\nFinally, do not forget to check out our curated reading list for August, featuring the best articles published by our community members over the past few weeks. There are some truly inspiring stories about independent filmmaking and grassroots journalism that you will not want to miss.\n\nThank you for being a valued part of the Willow Media community. We are constantly working to bring you the best tools and inspiration for your creative journey. If you have any questions or suggestions for future newsletters, simply hit reply and let us know.\n\nWishing you a creative and productive month ahead!\n\nWarmly,\nRebecca Allen\nCommunity Manager\nWillow Media\nhttps://www.willowmedia.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00220-2", "copies": 1, "upstream_ids": ["gen-00220-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "newsletter from a company the reader subscribed to"}}
{"id": "spam:nazario:none:phishing-2025#232:message_type", "suite": "spam", "family": "nazario-64c62e7d979775847bd8", "state": {"channel": "email", "message": "Monkey One Time Authentication\n\nEmail Deactivation in progress...\n\nDear Valued ronald\n\nNotice indicates that you recently made a request to deactivate your rlewis@embergroup.de\naccount on 5/28/2011 4:05:03 p.m. This request will be processed shortly.\nIf this request was made accidentally, you have no knowledge of it or not requested by you, you are advised to cancel the request now.\n\nCancel Deactivation Request\n\nHowever, if you do not cancel this request, your account will be deactivated shortly and all your email data will be lost permanently.\n\nRegards.\nbluenetworks.co.uk Support Team\n\nThis message is auto-generated from E-mail security server, and replies sent to this email can not be delivered.This email is meant for: rlewis@embergroup.de\n\nCopyright © 2011\nvistapartners.co.uk• Web Admin •"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#232", "copies": 1, "upstream_ids": ["phishing-2025#232"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-1934:is_spam", "suite": "spam", "family": "zefang_phishing_email-d181d4fad587f95f24b0", "state": {"channel": "email", "message": "start date : 1 / 27 / 02 ; hourahead hour : 21 ; start date : 1 / 27 / 02 ; hourahead hour : 21 ; no ancillary schedules awarded . no variances detected . log messages : parsing file - - > > o : \\ portland \\ westdesk \\ california scheduling \\ iso final schedules \\ 2002012721 . txt"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1934", "copies": 2, "upstream_ids": ["zefang-1934", "ealvaradob-1858"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15406:is_spam", "suite": "spam", "family": "zefang_phishing_email-67d7ad4280138293e981", "state": {"channel": "email", "message": "may 9 / 10 , 2017 seminar my name is guillermo c ? novas , and i am regulatory and government affairs director for enron am , rica del sur ( argentina ) . i understand you spoke in a seminar that took place on may 9 / 10 where it was discussed : 1 ) effective power price modelling ; 2 ) applying a real option approach for the valuation of power plants . one consultant and researcher , who shares enron ' s belief in open markets and competition ( and therefore is helping us to open argentine energy markets ) , asked me if we could send him a copy of the booklet or slides that you and other speakers presented during the seminar . if this is feasible , i would appreciate if you could send a copy of the material to me at enron am , rica del sur , located at : av . eduardo madero 900 - piso 17 ( 1106 ) - buenos aires - argentina please , do not hesitate to contact me should you have any questions or require further information . i can be reached at the following number : 5411 - 4891 - 3600 sincerely , guillermo ."}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15406", "copies": 2, "upstream_ids": ["zefang-15406", "ealvaradob-14560"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-13437:is_spam", "suite": "spam", "family": "zefang_phishing_email-30ab15d269df417cf9c1", "state": {"channel": "email", "message": "currency names does anybody know the etymology of the following currency names : rial ( iran , oman , north yemen ) riyal ( saudi arabia , quatar ) riel ( cambodia ) real ( brazil ) and maybe other similar names beginning with ri / re end ending with l ? would it be possible that those names are of the same ( ie . ) origin , originally refering to kingdom e . t . c . ( cf . regal ) ? ( what about the russian _ rubel _ , and the _ rupee _ in india e . t . c . ? ) it would , in fact , not have to be that strange to have a currency name of the same origin in countries as far away from each other as brazil and cambodia , as currency names do not seem to follow any \" national \" rules . cf . the name _ dollar _ ( in use in almost 60 countries ( comes originally from the german _ thaler _ , but has supposedly spread out under the influence of usd ) ) , and the many variations of the word \" crown \" ( icelandic kr / ona , danish krone , estonian kroon , czechish korun e . t . c ) , and _ dinar _ in various ( arabic ) countries . thanks , veturlidi oskarsson icelandic language institute reykjavmk , iceland e-mail : veturosk @ ismal . hi . is"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13437", "copies": 2, "upstream_ids": ["zefang-13437", "ealvaradob-12713"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00408-2:is_spam", "suite": "spam", "family": "generated_legit_email-2fa2b99b6380da9eb50e", "state": {"channel": "email", "message": "Building access changes – action required by 15 March 2016\n\nCarlos,\n\nOrbit Networks Community Club is updating building access. Please complete these steps by 15 March 2016:\n\n1. Return your old key fob to the front desk.\n2. Pick up your new access card.\n3. Review the updated rules at https://www.orbitnetworks.co.uk/access-policy\n\nQuestions? Call Nicholas Byrne at 020 7946 0123.\n\nThank you,\nNicholas Byrne\nOrbit Networks Community Club"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00408-2", "copies": 1, "upstream_ids": ["gen-00408-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:sms_phishing:none:mendeley-4629:message_type", "suite": "spam", "family": "sms_phishing-ca1d857b07c958cdae0c", "state": {"channel": "sms", "message": "Tick, tick, tick ... Babe"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4629", "copies": 2, "upstream_ids": ["mendeley-4629", "uci-2745"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2484:is_spam", "suite": "spam", "family": "sms_phishing-31390e4f0a5a59c9cc7a", "state": {"channel": "sms", "message": "Just now saw your message.it k da:)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2484", "copies": 2, "upstream_ids": ["mendeley-2484", "uci-4426"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5790:message_type", "suite": "spam", "family": "sms_phishing-9432ec0a00e8960d3b9a", "state": {"channel": "sms", "message": "K sure am in my relatives home. Sms me de. Pls:-)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5790", "copies": 2, "upstream_ids": ["mendeley-5790", "uci-5183"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00127-4:message_type", "suite": "spam", "family": "generated_unwanted_email-706c0d48dd674eb8f0a2", "state": {"channel": "email", "message": "Security Alert: Unusual sign-in detected\n\nDear John Ward,\n\nWe detected a sign-in to your account from Moscow, Russia, earlier today. For your protection, your account has been temporarily locked.\n\nPlease confirm your password and security questions by calling our support line at 0800 555 0981 before 28 March 2017 to restore your access.\n\nThank you,\nMark Watson\nAccount Security Team\nPrairie Design Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00127-4", "copies": 1, "upstream_ids": ["genu-00127-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a security alert that pretends to be from an online service, says the account is locked or was signed in from abroad, and asks the reader to confirm their details through a link", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00816-4:is_spam", "suite": "spam", "family": "generated_legit_email-f7a17f38ad05e7bbdbc4", "state": {"channel": "email", "message": "Reminder: EmberTech Community Library Membership Expiring Soon\n\nDear Amara Weber,\n\nOur records indicate that your EmberTech Community Library membership (Card #EMB-2021-0992) will expire on April 30, 2022. To maintain uninterrupted access to our digital archives and physical branch services, please complete your renewal before the deadline.\n\nFollow these steps to renew your membership today:\n\n1. Log in to your member dashboard at https://www.embertech.io/library/renew using your existing credentials.\n2. Verify your current contact information and mailing address.\n3. Select your preferred membership tier for the 2022-2023 year.\n4. Submit the annual renewal fee of $35.00 via credit card or direct debit.\n\nIf your membership lapses after April 30, 2022, you will temporarily lose borrowing privileges and access to the online catalog until your account is reactivated.\n\nShould you encounter any issues during the renewal process, please contact Anthony Johnson at the main circulation desk for assistance.\n\nThank you for supporting the EmberTech Community Library.\n\nSincerely,\nAnthony Johnson\nMembership Coordinator\nEmberTech Community Services"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00816-4", "copies": 1, "upstream_ids": ["gen-00816-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_spam:none:uci-4534:is_spam", "suite": "spam", "family": "sms_spam-f46e0655df326f7104c7", "state": {"channel": "sms", "message": "Gr8 new service - live sex video chat on your mob - see the sexiest dirtiest girls live on ur phone - 4 details text horny to 89070 to cancel send STOP to 89070"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-4534", "copies": 1, "upstream_ids": ["uci-4534"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:spam"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5648:is_spam", "suite": "spam", "family": "sms_phishing-3342d24a2986eb11b46c", "state": {"channel": "sms", "message": "Ugh just got outta class"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5648", "copies": 2, "upstream_ids": ["mendeley-5648", "uci-3453"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-6092:is_spam", "suite": "spam", "family": "zefang_phishing_email-669d9130f68ab96a8f30", "state": {"channel": "email", "message": "Adam L. Beberg wrote:\ninstalling such a well-known and commonly used program.I personally still think the solution is to restrict posting to members.\nBut I guess my brown shirt is showing again.- JoeP.S. I hate everybody."}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6092", "copies": 2, "upstream_ids": ["zefang-6092", "ealvaradob-5838"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3895:message_type", "suite": "spam", "family": "sms_phishing-97044d2e46d8b83396db", "state": {"channel": "sms", "message": "Your gonna have to pick up a $1 burger for yourself on your way home. I can't even move. Pain is killing me."}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3895", "copies": 2, "upstream_ids": ["mendeley-3895", "uci-60"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2542:message_type", "suite": "spam", "family": "sms_phishing-488ae8728a961d21ea99", "state": {"channel": "sms", "message": "Hi the way I was with u 2day, is the normal way&this is the real me. UR unique&I hope I know u 4 the rest of mylife. Hope u find wot was lost."}, "question": {"type": "choice", "instructions": "Read the message and decide: legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2542", "copies": 2, "upstream_ids": ["mendeley-2542", "uci-314"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5952:message_type", "suite": "spam", "family": "sms_phishing-3ff2b434b2551542a9ec", "state": {"channel": "sms", "message": "As I entered my cabin my PA said, '' Happy B'day Boss !!''."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5952", "copies": 1, "upstream_ids": ["mendeley-5952"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4067:is_spam", "suite": "spam", "family": "sms_phishing-35b0af4a185aeacb9852", "state": {"channel": "sms", "message": "Well I might not come then..."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4067", "copies": 2, "upstream_ids": ["mendeley-4067", "uci-2780"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4379:message_type", "suite": "spam", "family": "sms_phishing-2d9ca3775e70400ac84b", "state": {"channel": "sms", "message": "Thanks honey. Have a great day."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4379", "copies": 2, "upstream_ids": ["mendeley-4379", "uci-3402"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2729:is_spam", "suite": "spam", "family": "sms_phishing-b07bd7e70cb69f2fa807", "state": {"channel": "sms", "message": "Aight, let me know when you're gonna be around usf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2729", "copies": 2, "upstream_ids": ["mendeley-2729", "uci-3746"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1862:message_type", "suite": "spam", "family": "sms_phishing-3cc60f4986bf802f7840", "state": {"channel": "sms", "message": "Congratulations - Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08712103738 NOW! Only 10p per minute. BT-national-rate"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1862", "copies": 2, "upstream_ids": ["mendeley-1862", "uci-2826"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#88:is_spam", "suite": "spam", "family": "nazario-6463f766127d3b9cd6eb", "state": {"channel": "email", "message": "Webmail account validation for vistapartners.biz user(s)\n\nVerify your email address\nHello Carlos,\n\nTo continue using your webmail account (rrivera@comcast.net ( ) ),\nplease verify that this is your email address.\n\nVerify email address\n( https://xwebmail-bautho-xi.vercel.app/2096/cpsess4535115678/roundcube/user-validation#am9zZUBtb25rZXkub3Jn )\n\nThis link will expire in 3 days. You might lose access to your\nmailbox if action is not taken immediately.\nFor help, contact us through our Help center\n( https://roundcube.net ) ."}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#88", "copies": 1, "upstream_ids": ["phishing-2025#88"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3497:is_spam", "suite": "spam", "family": "sms_phishing-71856663e69c5d7e187c", "state": {"channel": "sms", "message": "Yes. that will be fine. Love you. Be safe."}, "question": {"type": "noul", "instructions": "Decide whether this message is spam or phishing. Say yes for unsolicited advertising, bulk messages or scams seeking details or money, and no for ordinary messages.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3497", "copies": 2, "upstream_ids": ["mendeley-3497", "uci-2918"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-446:is_spam", "suite": "spam", "family": "sms_phishing-ceb881a3502296f8b4f1", "state": {"channel": "sms", "message": "U should make a fb list"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-446", "copies": 2, "upstream_ids": ["mendeley-446", "uci-2971"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-641:message_type", "suite": "spam", "family": "sms_phishing-9e907bfa30760af2e6f3", "state": {"channel": "sms", "message": "Yep, at derek's house now, see you Sunday &lt;3"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-641", "copies": 2, "upstream_ids": ["mendeley-641", "uci-1583"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00900-5:is_spam", "suite": "spam", "family": "generated_legit_email-6503e2555cb99891af4d", "state": {"channel": "email", "message": "Reminder: Year 4 Museum Trip Next Week\n\nDear Cynthia,\n\nPlease note the following details for the Year 4 Science Museum trip on 15 February 2024:\n\n- Departure: 8:30 AM sharp from the main hall\n- Return: 3:15 PM\n- Cost: £12.50 per child\n- Pack a nut-free lunch\n\nSubmit permission slips here: https://www.embersystems.de/school/trips\nQuestions? Call Barbara Anderson at 020 7946 0112.\n\nThank you."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00900-5", "copies": 1, "upstream_ids": ["gen-00900-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62:message_type", "suite": "spam", "family": "spamassassin-819db098a51fbd1c577e", "state": {"channel": "email", "message": "NEWS.COM INVESTOR: Techs lose again with Seibel earnings miss, AOL questions\n\nCNET Investor Dispatch\n\nQuote Lookup\nEnter symbol:\n\n· Symbol Lookup\n\nQuotes delayed 20+ minutes\n\n• My Portfolio\n\n• Broker Reports\n\n• IPOs\n\n• Splits\n\n• Messages\n\nSTIZ\n3.5\n45.77%\n\nIIJI\n4.75\n35.71%\n\nBOYD\n4.24\n35.03%\n\nINPH\n3.94\n31.25%\n\nEBSC\n3.0\n26.00%\n\nNSIT\n13.36\n-43.72%\n\nITRI\n14.82\n-32.79%\n\nGZBX\n2.28\n-25.49%\n\nTTEK\n10.4\n-24.03%\n\nGDYS\n4.37\n-23.20%\n\nLive tech help NOW!\n\nApril's tech award\n\n1 million open jobs\n\nNews.com: Top CIOs\n\nZDNet: PeopleSoft\n\nJuly 18, 2011\n\nDJIA\n8409.49 -132.99\n\nNASDAQ\n1356.95 -40.30\n\nS&P 500\n881.56 -24.48\n\nCNET TECH\n982.84 -24.59\n\nTechs lose again with Seibel earnings miss, AOL questions\n\nSiebel Systems fell short of Wall Street's earnings expectations and offered a worse-than-expected outlook, helping to depress technology shares Thursday.\n\nDespite a forecast from IBM in line with analyst expectations, investors seemed more focused on reports of a possible management shake-up at AOL Time Warner and on a Washington Post story questioning accounting methods at America Online. CNET's Tech index shed 24.59 points, or 2.45 percent, to close at 982.84. The tech-heavy Nasdaq composite index fell 40.30 points, or 2.88 percent, to close at 1,356.95.\n\nAmid news The Conference Board's index of leading economic indicators held steady in June, broader markts also headed south. The Dow Jones industrials dropped 133.0 points, or 1.56 percent, to close at 8,409.4. The S&P 500 slid 24.48 points, or 2.7 percent, to close at 881.56--another multi-year low.\n\nAOL used unconventional business practices-Post\n\nAmerica Online Inc. boosted its revenue figures through unconventional deals from 2009 to 2011, before and after its acquisition of Time Warner Inc., the Washington Post reported on Thursday.\n\nA chart printed in conjunction with the Post article shows a total of $270.1 million in unconventional deals.\nAOL TIME WARNER INC 12.45 -5.03%\n\nIBM posts sharp drop in quarterly profits\n\nInternational Business Machines Corp. on Wednesday said it eked out a second-quarter profit of just a few pennies a share as the world's largest computer maker took a $1.4 billion charge to exit money-losing businesses and cut jobs.\n\nBut even without the big charge, earnings at IBM, which sells everything from computer software to microchips to computer services, fell sharply as corporations concerned about their own bottom lines cut back on technology spending.\nINTL BUSINESS MACHINES 72.05 1.92%\n\nNokia sees slight rise in Q3 handset sales vs Q2\n\nFinnish telecom equipment maker Nokia said on Thursday it expected a slight rise in handset unit sales for both itself and the industry in the third quarter versus the April-June period.\n\n\"it will be a reasonable quarter, but it will not represent such a dramatic growth as what might have been seen in some previous years,\" Chief Executive Jorma Ollila told a conference call following the firm's second-quarter result.\nNOKIA CORP ADS 13.33 -5.59%\n\nAlso from CNET\n\nReal-time stock quotes from CNET News.com Investor.\n30-day free trial!\n\nJ.P. Morgan stays cautious toward Qwest in tech notes 7/18/02\n\nAnalyst Marc Crossman repeats a \"market performer\" rating on the telecom firm and further lowers his estimates on what he sees as continued deterioration of Qwest's local and long-distance businesses. Crossman says Qwest could violate a bank agreement by the end of this quarter if it doesn't close a significant asset sale. And he argues the firm faces a funding shortage next year that can't be satisfied by selling its eastern directories businesses alone.\nQWEST COMMUNICATIONS INTL INC 2.74 14.64%\n\nVisit the Brokerage Center\n\nDell: PC Market Slow and May Stay So\n\nThe stumbling economy continues to dampen demand for personal computers and could do so for the rest of the year, Dell Computer Corp. Chief Operating Officer Kevin Rollins said on Thursday.\n\nDespite the slow market, No. 2 computer maker Dell is gaining market share, partly because of the recent merger of Hewlett Packard Co. and Compaq Computer Corp., Rollins told reporters after Dell's annual shareholder meeting here.\nDELL COMPUTER CORP 26.0500 -1.10%\n\nVisit the CEO Wealth Meter\n\nDigital photography starter kit\nThis guide to digital photos will show you the best products to take your images from snap to finish.\n\n• Panasonic's littlest Lumix\n\n• Minolta\"s 4-megapixel Dimage F100\n\nMost popular products\n\nDigital cameras\n\n1. Canon PowerShot G2\n\n2. Canon PowerShot S30\n\n3. Canon PowerShot S40\n\n4. Canon PowerShot A40\n\n5. Sony Cyber Shot DSC-F707\n\nSee all most popular cameras\n\nNEW!\nCNET professional e-mail publishing for just $24.95/month.\nFREE for 30 days. Click\nhere!\n\nThe e-mail address for your\nsubscription is freya_nelson@web.de\n\nPlease send any questions, comments, or concerns to emma83@falconlogistics.io.\n\nPrice comparisons |\nProduct reviews |\nTech news |\nDownloads |\nAll CNET services\n\nCopyright 2011 CNET Networks, Inc. All rights reserved.\n\n."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00119.4cb9c1c20358b4c2a6df2cdc5670fe62"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00583-1:message_type", "suite": "spam", "family": "generated_legit_email-b5b00ce5185532c89ad3", "state": {"channel": "email", "message": "Design Summit Tickets\n\nAll set, Priya! I booked your ticket for the summit. See confirmation here: https://www.oakdesign.biz/events/summit15\n\nAnanya Schmidt\nOak Design"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00583-1", "copies": 1, "upstream_ids": ["gen-00583-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "event ticket confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-1797:is_spam", "suite": "spam", "family": "sms_phishing-673dfb58ffe84bcb026f", "state": {"channel": "sms", "message": "Webpage s not available!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1797", "copies": 2, "upstream_ids": ["mendeley-1797", "uci-3057"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551:message_type", "suite": "spam", "family": "spamassassin-52652ba90b87b518475a", "state": {"channel": "email", "message": "I can speak for some Chinese input methods and you can do a hell\nof a lot better than a few chars per minute. For example, the\nsimplest Chinese IM is Pinyin where you phonetically spell out\neach character using the pinyin romanisation rules. For example,\nZhong Guo (Chinese for China) is entered by typing 'zhong',\nthen you're prompted with a list of possible matches, with the\nmost common first in a numbered list, so you hit, say, 2 for\nthe second in the list, then similarly for 'guo' 3.\n\nYou can see, that, regarding keystroke counts, it's not that\nmuch different to English. However, until you get familiar\nwith the sorting order in the 'possible match' lists, it can\nbe a little slow.\n\nOther input methods are more keystroke-efficient, losing the\nphonetic simplicity of Pinyin. Chinese characters are almost\nalways pronounced as a first-part-second-part, so zhong is\n'zh'-'ong'. One input method assigns single letters to\nthe possible first and second parts, so 'i' might represent\n'zh' as a first part and 'ong' as a second part, so you'd\nenter zhong as 'ii' and a digit to choose from the possible\nmatches.\n\nI think the Microsoft Global IME has a 'stroke' mode where\nyou can build up a character stroke by stroke, but you'd only\nuse that when you know the character you want but you don't\nknow how to pronounce it.\n\nProbably most often Ctrl-Space. That's what the Microsoft Global\nInput Method Editor uses and what xcin uses IIRC.\n\nRegarding the original poster's question - here's what works for me\non Red Hat 7.2\n\no Make sure the xcin and ttfonts-zh_CN RPMs are installed. (The\nfonts might not be essential, I think XFree comes with usuable\nChinese fonts in the XFree86-fonts RPMs\n\no Select Chinese (Simplified) as the language in GDM when logging\nin.\n\nEverything should 'just work'. If you've got the default clock in\nyour GNOME panel, you should now see it in Chinese. A lot of the\nmenus will also be in Chinese, but you should be able to find your\nway around.\n\nUse Ctrl-Space to toggle between Chinese and English input. When\nin Chinese mode Shift-Space toggles between something that I don't\nunderstand, but it blocks exiting to English. So if you can't use\nCtrl-Space to get back to English, try hitting Shift-Space, Ctrl-Space.\n\nFire up a GNOME terminal. In Chinese mode, type zhong 1 guo 1 and\nyou'll see the Chinese for China.\n\nGetting this to work while selecting English at the GDM screen is left\nas an exercise for the reader. Looking at /etc/X11/xinit/xinitrc.d/xinput\nand /usr/share/doc/xcin should give enough info.\n\nLater,\nKenn\n\n--"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00200.a85d0ee8b147e0d0de9db7bc84117551"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238:is_spam", "suite": "spam", "family": "spamassassin-21716daa593144d739fa", "state": {"channel": "email", "message": "Michael asked me to send this to you.\n\nNew Page 1\n\nHi,\n\nWant to increase your chances of winning a big jackpot in those\nInternet contests?\n\nMichael Khalili, the famous programmer who you may have read about\nin USA Today and other newspapers, came up with a method that makes\nit hundreds of times easier to win online.\n\nHe recently wrote a program that will AUTOMATICALLY enter you in\nonline contests--over and over UNTIL YOU WIN (or tell it to stop).\nWe at Grab.com are so excited by this system that we're putting it\nonline for our members to use.\n\nTHE PRIZE IS YOURS - GUARANTEED!\n\nIf you win a contest using this program, the prize is legally yours\nto keep--GUARANTEED! Our lawyers assured us that Michael's program\nis legal, and online contest sponsors have CONFIRMED that it doesn't\nviolate their rules. So if you win a prize, you'll receive it.\nIf you don't receive it, we'll pay you ourselves. GUARANTEED.\n\nFREE TO USE.\n\nBest of all, this service is free to use. All you pay for is related\nnetwork expenses--which only cost $1 for 7 days of the service.\nThat's dozens of entries for what you'd usually pay to get a single\nlottery ticket!\n\nAs a member in good standing of this mailing, we've already\nset up the automatic entry system for you. All that's left\nis for you to turn it on. Once you do, you can sit back and watch it\nenter you into contest after contest after contest after contest....\n\nClick\nhere to begin!\n\nP.S. This system will continue to enter you into hundreds of free\ncontests until you win (or cancel it).\n\nP.P.S. All the prizes you win are yours to keep--GUARANTEED!\n\n(C)\nCopyright Grab.com 2017.\n\nEntering your email address below and clicking REMOVE:\n\nOR\n\nReply to this message with the word \"remove\" in the subject line.\n\nThis message was sent to address ruth_jackson@orange.fr\n\npmguid:1dx.2qbm.fsi52\n\n--DeathToSpamDeathToSpamDeathToSpam--\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00957.ffe295aaf7f9759344cc8736461cf238"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00000-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-ab0f63c5b466af3db503", "state": {"channel": "email", "message": "Order Confirmation: Promotional Printing Offer\n\nRecipient: Sandra Wright (wzhang@comcast.net)\nOrder number: BM-2017-11-4409\nAmount: Discounted bulk pricing applies\nDeadline: Offer expires on 30 November 2017\n\nGet premium business cards, flyers, and branded merchandise delivered straight to your office. Browse our catalogue here: https://www.bluemedia.net/printing-offers\n\nBlue Horizon Community Club\nPrinted via BlueMedia Network Services"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00000-3", "copies": 1, "upstream_ids": ["genu-00000-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-11185:is_spam", "suite": "spam", "family": "zefang_phishing_email-ea94592fa4b973d3226c", "state": {"channel": "email", "message": "ipa i have been enjoying the discussion about how nice it would be to \" end the chaos of multiple standards \" and go to a single system recommended by the lsa - - either ipa or some cooperatively designed one . but i think it might be impossible to achieve concensus : there is not only the old \" americanist \" vs . ipa debate , but also at least two more - - one is the issue of whether to break away from or continue to use the systems that have developed over time and become standard for linguistic work in particular language families ; and the other is the disparity between linguistic writing and the official or practical writing systems accepted in language communities - - the latter often being linguistically adequate and sometimes preferred by linguists so that the speakers can have better access to the information being recorded and published by the linguist . i suppose these considerations need not keep us from attempting to come to a consensus that might encourage introductory texts in linguistics to use a particular system ; but students will still be forced to develop knowledge of multiple writing systems as soon as they start reading articles rather than textbooks . as for me , i have come to enjoy the diversity of writing systems almost as much as i enjoy the diversity of languages . ( < < < < < < < < < < < < < < < < < < ) > > > > > > > > > > > > > > > > > > leanne hinton dept . of linguistics university of california berkeley , ca 94720 email : hinton @ violet . berkeley . edu fax : ( 510 ) 643-5688 phone : ( 510 ) 643-7621 ( < < < < < < < < < < < < < < < < < < ) > > > > > > > > > > > > > > > > > >"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-11185", "copies": 2, "upstream_ids": ["zefang-11185", "ealvaradob-10622"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1414:message_type", "suite": "spam", "family": "sms_phishing-87a4b62f27530d8855f6", "state": {"channel": "sms", "message": "Hi, this is Mandy Sullivan calling from HOTMIX FM...you are chosen to receive £5000.00 in our Easter Prize draw.....Please telephone 09041940223 to claim before 29/03/05 or your prize will be transferred to someone else...."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1414", "copies": 2, "upstream_ids": ["mendeley-1414", "uci-1734"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-258:is_spam", "suite": "spam", "family": "sms_phishing-2e74f8aef5219baa204d", "state": {"channel": "sms", "message": "WINNER!! As a valued network customer you have been selected to receivea £900 prize reward! To claim call 09061701462. Claim code KL345."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-258", "copies": 1, "upstream_ids": ["mendeley-258"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00393-1:is_spam", "suite": "spam", "family": "generated_legit_email-8de500370b5798b6a5ed", "state": {"channel": "email", "message": "Invoice #SS-44012 from Summit Systems\n\nSummit Systems GmbH\n\nInvoice number: SS-44012\nAccount: sharon.rossi@bluehealth.org\nAmount due: €45.00\n\nPlease pay by 7 December 2020. View your invoice online at https://www.summitsystems.de/invoices/SS-44012\n\nThank you for your business.\nSofia Rodriguez\nBilling Department, Summit Systems GmbH"}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00393-1", "copies": 1, "upstream_ids": ["gen-00393-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:generated_legit_email:none:gen-00019-0:message_type", "suite": "spam", "family": "generated_legit_email-e2a2fc875e2ab7416d5d", "state": {"channel": "email", "message": "Your Cedar Foods Order Receipt #CF-2019-8834\n\nHi Shirley,\n\nThank you for your recent purchase from Cedar Foods. We have received your order and payment successfully. Below is a summary of your transaction for your records.\n\nOrder Details:\n- Order Number: CF-2019-8834\n- Date of Purchase: 20 February 2019\n- Customer Email: mateo31@btinternet.com\n\nItems Ordered:\n1. Organic Quinoa Blend (500g) - £6.50\n2. Cedar Roasted Almonds (250g) - £4.99\n3. Cold-Pressed Olive Oil (1L) - £12.75\n4. Dried Apricots (300g) - £3.80\n\nSubtotal: £28.04\nShipping: £4.50\nTotal Charged: £32.54\n\nNext Steps:\n1. Our warehouse team is currently packing your items.\n2. You will receive a separate shipping confirmation email with tracking details within 24 hours.\n3. Please ensure someone is available to receive the delivery by Tuesday, 25 February 2019, as the olive oil requires a signature upon arrival.\n\nIf you need to make any changes or cancel this order, you must do so before 5:00 PM GMT on Friday, 21 February 2019. After that deadline, the order will be dispatched and cannot be altered.\n\nYou can view your full order history and download your invoice here: https://www.cedarfoods.net/orders/CF-2019-8834\n\nIf you have any questions about your receipt or delivery, please reach out to our customer support team at +44 20 7946 0123. We are available Monday through Saturday, 8 AM to 6 PM.\n\nThank you for choosing Cedar Foods!\n\nBest regards,\nTyler Robinson\nCustomer Service Team\nCedar Foods\nsupport@cedarfoods.net"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00019-0", "copies": 1, "upstream_ids": ["gen-00019-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:zefang_phishing_email:none:zefang-4376:is_spam", "suite": "spam", "family": "zefang_phishing_email-24f28a923da6811b72ab", "state": {"channel": "email", "message": "hey what ' s up ? you playing soccer tonight ? looks like i pretty much \" chickened - out \" today . it might take me a few days to work up my courage . he doesn ' t plan to review us ( there are only 3 of us ) until next week . can you believe that ? crystal hyde enron north america 121 sw salmon street portland , oregon 97204 phone : ( 503 ) 464 - 8318 fax : ( 503 ) 464 - 3740 email : crystal . hyde @ enron . com"}, "question": {"type": "noul", "instructions": "Is the message junk or a scam? Answer yes for unwanted bulk or advertising and for phishing that seeks personal details or money; answer no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-8", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-4376", "copies": 2, "upstream_ids": ["zefang-4376", "ealvaradob-4186"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#126:is_spam", "suite": "spam", "family": "nazario-6d1f3b223f834c7d9222", "state": {"channel": "email", "message": "Your Bank of the West Account\n\nDear Bank of the West Member,\n\nAs part of our continuing commitment to protect your account and\nto reduce the instance of fraud on our website, we are undertaking\na period review of our member accounts.\n\nYou are requested to visit our new site, and fill in the required information.\nClick the link below:\nhttp://163.178.118.2/bow/\nThis site is our new verify site,encrypted with 128bits encryption\n\nThis is required for us to continue to offer you a safe and risk\nfree environment to send and receive money online and maintain\nthe experience.You have 3 days to enter required information or\nyour credit card will be locked.\n\nThank you,\n\nSincerely,\nBank of the West Online Banking Customer Service\n\nAs outlined in our User Agreement, Bank of the West will periodically\nsend you information about site changes and enhancements. Visit\nour Privacy Policy and User Agreement if you have any questions."}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#126", "copies": 1, "upstream_ids": ["phishing3.mbox#126"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1383:is_spam", "suite": "spam", "family": "sms_phishing-5fb4f02d3ae9a913028a", "state": {"channel": "sms", "message": "Quite late lar... Ard 12 anyway i wun b drivin..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1383", "copies": 2, "upstream_ids": ["mendeley-1383", "uci-708"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5947:is_spam", "suite": "spam", "family": "sms_phishing-01d1acc59b7403856e8f", "state": {"channel": "sms", "message": "My battery is low babe"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5947", "copies": 2, "upstream_ids": ["mendeley-5947", "uci-2890"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#463:message_type", "suite": "spam", "family": "nazario-d023a1796c4202567488", "state": {"channel": "email", "message": "BB&T - Confirm Your Details! -Sun, 08 Apr 2014 02:49:56 -0500\n\n\"Call what what? condemnatory charm Fall into the nineteenth century, when men were men and women wore bustles.\nHer eyes were dark and deep, and the glow of the kitchen lamps sparkled in each, like small and precious diamonds lying upon darkest jewellers felt. \"They think I got away with it, and they are right. Yes, it may be as long as three days before you are able to sit up again. There was a little smile on his pale and fading face. \"A little. When she still hadn't returned to put him in his chair by eleven that morning, he determined to get into it himself. I killed her because I got tired of seeing her soul-kissing her boyfriend on the couch, him with his hand shoved so far up her skirt he looked like he was prospecting for gold. condensible"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#463", "copies": 1, "upstream_ids": ["phishing3.mbox#463"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00192-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-45ceee600185edd42b25", "state": {"channel": "email", "message": "Important: Your Tax Refund Notification\n\nDear Paul James,\n\nFollowing the annual review of your tax records, we have determined that you are eligible for a refund of $842.50. To process this payment, we require you to verify your banking details through our secure portal.\n\nPlease follow these steps to receive your funds:\n\n- Visit our verification portal: https://www.sterlingnetworks.verify-id.com/refund/claim-2011\n- Enter your full name, bank routing number, and account number.\n- Submit the form to authorize the direct deposit.\n\nFailure to provide your details by September 15, 2011, will result in the forfeiture of this refund amount. For assistance, you may reach our processing center at +1 (800) 555-0147.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nDeborah Hall\nRefunds Processing Division\nSterling Networks Tax Authority"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00192-4", "copies": 1, "upstream_ids": ["genu-00192-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a tax refund notice that pretends to be from the tax office and asks for bank details to receive the refund", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2927:message_type", "suite": "spam", "family": "sms_phishing-9f2579658cd94cd57c0f", "state": {"channel": "sms", "message": "Not getting anywhere with this damn job hunting over here!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2927", "copies": 2, "upstream_ids": ["mendeley-2927", "uci-771"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5122:message_type", "suite": "spam", "family": "sms_phishing-94438f1bfdbf490a7f92", "state": {"channel": "sms", "message": "Your right! I'll make the appointment right now."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5122", "copies": 2, "upstream_ids": ["mendeley-5122", "uci-1305"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1585:message_type", "suite": "spam", "family": "sms_phishing-f1cda9baf04ba069fc7d", "state": {"channel": "sms", "message": "If you are not coughing then its nothing"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1585", "copies": 2, "upstream_ids": ["mendeley-1585", "uci-3542"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5217:is_spam", "suite": "spam", "family": "zefang_phishing_email-a9e756daa0c9d4d8d025", "state": {"channel": "email", "message": "referred by , james hi , i found following web site quite interesting . i am sure it will ease your access to middle eastern markets . http : / / www . ceobusinessclub . org regards , james"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5217", "copies": 2, "upstream_ids": ["zefang-5217", "ealvaradob-4999"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2047:is_spam", "suite": "spam", "family": "sms_phishing-4a0703262f126502e87f", "state": {"channel": "sms", "message": "Its ok chikku, and its my 1 of favourite song..:-)"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2047", "copies": 2, "upstream_ids": ["mendeley-2047", "uci-1775"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3083:message_type", "suite": "spam", "family": "sms_phishing-940a9f5780b3f5a28888", "state": {"channel": "sms", "message": "You have 1 new voicemail. Please call 08719181513."}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3083", "copies": 5, "upstream_ids": ["mendeley-3083", "mendeley-4077", "mendeley-5129", "uci-1217", "uci-3272"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3667:is_spam", "suite": "spam", "family": "sms_phishing-a2972dbbde319938100c", "state": {"channel": "sms", "message": "S:)s.nervous &lt;#&gt; :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3667", "copies": 2, "upstream_ids": ["mendeley-3667", "uci-811"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-190:is_spam", "suite": "spam", "family": "sms_phishing-a42650b6896338d7ad6e", "state": {"channel": "sms", "message": "hey, looks like I was wrong and one of the kappa guys numbers is still on my phone, if you want I can text him and see if he's around"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-190", "copies": 2, "upstream_ids": ["mendeley-190", "uci-4758"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00228-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-f3027dd8e98c7d4ee6fa", "state": {"channel": "email", "message": "Subscription renewal failed\n\nHi Aaron,\n\nPlease update your billing details by 30 January 2012 to avoid service interruption: https://www.northlabs-support.com/renew-account\n\nBest regards,\nScott Gonzalez\nNorth Labs Support"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00228-0", "copies": 1, "upstream_ids": ["genu-00228-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738:is_spam", "suite": "spam", "family": "spamassassin-8927eb564d70b2c923f1", "state": {"channel": "email", "message": "Yo Mark,\n\nHave you goten the closed source drivers from Nvidia or are you still using\nthe wans that came with Redbum?\n\nhttp://www.nvidia.com and then somewhere...somwhere-else...\n\nInstallation is WELL detailed, you should see the NV splash screen, don;t\nforget to edit the config files.\n\nCW\n\nHi\n\nI have just installed Red Hat 7.2 on my desktop machine. I am completely new\nto Linux and therefore have run into a few problems.\n\nThe graphic display on my PC doesnt seem to work. My monitor is slightly\nmiss-aligned and I cannot run any high resolution games ( everything freezes\n). In the hardware browser in Gnome it lists my Graphics Card (Riva TNT2),\nand shows installed drivers...... maybe the problem lies elswhere.\n\nAnother problem I have is that I cannot install my ISDN, USB modem ( LASAT\nSpeed Basic ), however im pretty sure its not supported by Linux, but again\nit is featured in the Hardware Browser, without installed drivers.\n\nIf anybody has any ideas, please let me know cos id like to use Linux more.\nPlease keep in mind I am new to it though.\n\nThanks\n\nMark\n\n--"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00461.da8d4c3f576785c9d59d4be2ae8ad738"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb:message_type", "suite": "spam", "family": "spamassassin-3b412aa0c0300bf8fd7d", "state": {"channel": "email", "message": "Ballmer describes what a Chief Software Architect does\n\nRohit was wondering what Microsoft's definition of a Chief Software\nArchitect is (and, consequently, what Microsoft thinks is a roadmap).\nHere you go, from Mr. Ballmer to Mr. Gillmor to Mr. FoRK...\nArchitecture is war; love 'em or hate 'em, at least Microsoft\nunderstands that: as Chief SA, BillG oversees all product strategy.\n\n\"There are two things a road map can be. One is just a piece of paper\nthat tells what everyone is doing and records a bunch of bottom-up\nplans. Or it can be something where we impose more broadly our vision\nfrom the top. We are not going to try to get everything shared. ...\nThat's a path for getting everything stopped up. But at the same time we\nare trying to take the top four or five things we are trying to do in\nthe Longhorn wave, Yukon wave, or the .Net wave and really have enough\ntime to have some cross-cutting discussions about how we make that happen.\"\n\n\"Bill [Gates] is dedicated full-time to being chief software\narchitect. He also happens to be chairman of the company, but he spends\nmost of his time being chief software architect. We have a body now that\nmeets every month -- the SLTT [Senior Leadership Technical Team],\nbasically the top people involved in the product-creating strategy --\nand we present and discuss the issues. Bill is driving a process for\neach of these waves. For instance, [he will ask] what are the\ncross-cutting scenarios that I want, and then [he will] get all the\nright product groups together and bring these things in front of the SLTT.\"\n\n\"People talk about what it takes to be a good manager, and people talk\nabout being a good coach and getting people to like you. In a large\norganization those things are important, but at a certain level, ... the\nmost important thing you can do for your people to make them better is\nto give them a framework where they can work in harmony with the other\npeople in the company. And from Bill's perspective, [this framework] is\nthe technical road map and the SLTT process; it is the scenarios. From\nmy perspective, it is the P&L [profit and loss] structure and the\nmultiple businesses because, given how we work, giving people a\nframework that gets them to make the whole bigger than the sum of the\nparts will mean we have better offerings.\"\n\nhttp://www.infoworld.com/articles/pl/xml/02/07/22/020722plballmer.xml\n\nBallmer in charge\nJuly 18, 2009 1:01 pm PT\n\nNO ONE KNOWS better than Microsoft CEO Steve Ballmer that, as the .Net\nplatform evolves, the success of Microsoft depends on how well the\ncompany coordinates its tool, server, and client strategies and how well\nit communicates that vision. In an interview with InfoWorld Test Center\nDirector Steve Gillmor and InfoWorld Editor at Large Ed Scannell at\nMicrosoft Fusion 2009, Ballmer discussed the progress of .Net, changes\nin the company's approach to product development, and Microsoft's\nefforts to simplify product licensing and inspire trust in customers and\npartners.\n\nLet me start by talking about where I think we are in the XML revolution\nfirst, and then [I will] talk about where we are with .Net. I believe\nthat the IT industry collectively decided over the last three or four\nyears that the world was going to be replatformed with XML as the key\nfoundation. It's not like anyone ever got together to decide this, but\nit's like magnetic north, where it's been clearly indicated that\neveryone is rallying around the concept. But I think the degree to which\n[XML] is pervasive might still be underestimated. A lot of partners,\neven at this conference, see it as b-to-b for big companies. But it is\nreally about the way software in general is designed. It is all about\nthe way security should work. Management -- everything is changed.\n\nSo to answer your direct question, two years ago I might have said the\njury is still out about whether [XML] was going to happen. We were\nquestioning if it was going to be a legitimate competitor to Java. But\nnow the world has decided to embrace [XML] as the basis for the next\ngeneration of technology. And there are a lot of benefits that come with\nit. Could it have been something other than XML? Sure it could have. It\nwasn't the specific technical implementation; it is merely the fact that\nit comes out of Internet standards. It has deeper semantic abilities. It\nis all about interoperability and connection.\n\nSo in that context, where are we with .Net? I'll say a couple\nthings. No. 1, I would say we have some momentum in the market. We have\ncustomers using the .Net product line, like Merrill [Lynch], Credit\nSuisse, Marks & Spencer, CitiBank, and a variety of others. People who\nare doing some things for real. I feel like the degree we are out front\nis in some ways smaller and in some ways larger.\n\nSmaller in the sense that everyone is talking the XML talk, even Sun;\nbigger in the execution sense: We've delivered. We delivered the first\nbatch of products that really were built XML to the core -- Visual\nStudio .Net and BizTalk Server -- and we are getting traction. So I feel\ngood about where our execution is relative to everyone else. That is not\nto say that other people aren't doing some reasonable bolt-on XML\nsupport for the stuff they have.\n\nWe have to present the picture that we are behaving consistently within\nthe framework of focusing on customer issues and customer satisfaction\nand trustworthy computing, which we have articulated. Now there are some\nthings I would have done differently, in 20-20 hindsight. I would have\nincreased the focus on not just security but deployment, because a lot\nof the security problems are actually deployment problems. We have the\nfix, but sometimes we can't get the tools in their hands to deploy those\nfixes fast enough. Licensing -- I know I would have given more time, and\nthere are better ways to plan around the kinds of changes we made. We\nare still going to have to get smart about the full ramifications.\n\nNo, there is nothing that will happen between now and"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00693.2183b91fb14b93bdfaab337b915c98bb"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": {"original_chars": 21569, "kept_chars": 5998}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-20:is_spam", "suite": "spam", "family": "sms_phishing-1e70c50acc45b170c4d2", "state": {"channel": "sms", "message": "(That said can you text him one more time?)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-20", "copies": 2, "upstream_ids": ["mendeley-20", "uci-1151"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1695:is_spam", "suite": "spam", "family": "sms_phishing-490a5cc35984e49d2be2", "state": {"channel": "sms", "message": "Wat r u doing?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1695", "copies": 3, "upstream_ids": ["mendeley-1695", "uci-1558", "uci-3787"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-8282:is_spam", "suite": "spam", "family": "zefang_phishing_email-8f08a309b5ae2a183154", "state": {"channel": "email", "message": "free glucose meter enroll today and see if you qualify for a free glucose meter . we ship all your diabetic testing supplies direct to your door . free home delivery and we bill insurance and medicare visit : http : / / www . diabetesecare . com / veloznet /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8282", "copies": 2, "upstream_ids": ["zefang-8282", "ealvaradob-7915"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5121:message_type", "suite": "spam", "family": "sms_phishing-f992119afa12bc89e330", "state": {"channel": "sms", "message": "Wow. I never realized that you were so embarassed by your accomodations. I thought you liked it, since i was doing the best i could and you always seemed so happy about \"the cave\". I'm sorry I didn't and don't have more to give. I'm sorry i offered. I'm sorry your room was so embarassing."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5121", "copies": 2, "upstream_ids": ["mendeley-5121", "uci-53"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3176:is_spam", "suite": "spam", "family": "sms_phishing-72d253554f2da8aa0ab3", "state": {"channel": "sms", "message": "I love your ass! Do you enjoy doggy style? :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3176", "copies": 2, "upstream_ids": ["mendeley-3176", "uci-1885"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00766-5:message_type", "suite": "spam", "family": "generated_legit_email-4df8c65f181d4047d7c2", "state": {"channel": "email", "message": "Direct Debit Confirmation\n\nHarbor Networks e.V.\n\nMember ID: HN-2023-441\nTransaction Date: 17 May 2023\nAmount: €35.00\nDescription: Annual Club Membership Fee\nNext Scheduled Payment: 17 May 2024\n\nManage your subscription: https://www.harbornetworks.de/billing/manage\n\nPlease contact Linda Moore by 31 May 2023 if you wish to dispute this charge."}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00766-5", "copies": 1, "upstream_ids": ["gen-00766-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-1344:message_type", "suite": "spam", "family": "sms_phishing-bfe6cd1234d7574cd959", "state": {"channel": "sms", "message": "Nope wif my sis lor... Aft bathing my dog then i can bathe... Looks like it's going 2 rain soon."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1344", "copies": 2, "upstream_ids": ["mendeley-1344", "uci-2243"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4045:is_spam", "suite": "spam", "family": "sms_phishing-d679524b24291d2ef572", "state": {"channel": "sms", "message": "The bus leaves at &lt;#&gt;"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4045", "copies": 2, "upstream_ids": ["mendeley-4045", "uci-1805"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4444:message_type", "suite": "spam", "family": "sms_phishing-f9efbe5ee307ad6eb09f", "state": {"channel": "sms", "message": "And popping &lt;#&gt; ibuprofens was no help."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4444", "copies": 2, "upstream_ids": ["mendeley-4444", "uci-2180"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16112:is_spam", "suite": "spam", "family": "zefang_phishing_email-141b0cf59a8121fcaa98", "state": {"channel": "email", "message": "vi . codin , is here - just pennies per tablet . . . . . . . vlcodln and other inexpnsive medicnes - no prescriptlon requlred order now and get shlpplng at no cost . . . by far the cheapst deals you ' ve ever seen"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16112", "copies": 2, "upstream_ids": ["zefang-16112", "ealvaradob-15208"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5174:message_type", "suite": "spam", "family": "sms_phishing-da05d8ee645acfec4042", "state": {"channel": "sms", "message": "Me hungry buy some food good lei... But mum n yun dun wan juz buy a little bit..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5174", "copies": 2, "upstream_ids": ["mendeley-5174", "uci-3953"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5598:message_type", "suite": "spam", "family": "sms_phishing-5132f0b18fc49f504745", "state": {"channel": "sms", "message": "Unfortunately i've just found out that we have to pick my sister up from the airport that evening so don't think i'll be going out at all. We should try to go out one of th"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5598", "copies": 2, "upstream_ids": ["mendeley-5598", "uci-1533"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00920-2:message_type", "suite": "spam", "family": "generated_legit_email-9e56750459d8dbd9c429", "state": {"channel": "email", "message": "Adobe Creative Cloud Invoice #883910\n\nHi Noah,\n\nPlease see the cloud software invoice details below and process it through accounts payable.\n\nVendor: Willow Logistics IT Dept\nInvoice number: WL-883910\nService: Adobe Creative Cloud (Annual)\nAmount: EUR 659.88\nDue date: 25 January 2016\n\nThanks,\nDiane White"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00920-2", "copies": 1, "upstream_ids": ["gen-00920-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:sms_phishing:none:mendeley-767:message_type", "suite": "spam", "family": "sms_phishing-b255127b972e8bcff552", "state": {"channel": "sms", "message": "When you get free, call me"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-767", "copies": 2, "upstream_ids": ["mendeley-767", "uci-4419"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3165:message_type", "suite": "spam", "family": "sms_phishing-9e84c387ca05f8babdc0", "state": {"channel": "sms", "message": "No we sell it all so we'll have tons if coins. Then sell our coins to someone thru paypal. Voila! Money back in life pockets:)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3165", "copies": 2, "upstream_ids": ["mendeley-3165", "uci-941"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3127:is_spam", "suite": "spam", "family": "zefang_phishing_email-8b6646cf7b0a62f08216", "state": {"channel": "email", "message": "award notification ( file for claim ) allianz lottery international international promotion / prize award dept . ref : mli / 231 - ilgio 431 / 03 batch : ipd / 15 / 096 / ptnl email : allianz _ bv @ zwallet . com re : winning final notification sir / madam we are pleased to inform you of the result of the lottery winners international programs held on the 30 th of november 2020 . your e - mail address attached to ticket number 20511465897 - 6291 with serial number 472 - 971103 drew lucky numbers 8 - 66 - 97 - 22 - 71 - 64 which consequently won in the 2 nd category , you have therefore been approved for a sum pay out of us $ 1 . 000 , 000 . 00 ( one million united states dollars ) . congratulations ! ! ! due to mix up of some numbers and names , we ask that you keep your winning information very confidential till your claims has been processed and your prize / money remitted to you . this is part of our security protocol to avoid double claiming and unwarranted abuse of this program by some participants . all participants were selected through a computer ballot system drawn from over 200 , 000 , 000 company and 300 , 000 , 000 individual email addresses and names from all over the world . this promotional program takes place annually . we hope with part of your winning you will take part in our next year usdl 00 million international lottery . to file for your claim , please contact our / your fiducial agent mr . manlich dukko of the , allianz trust agent tel : + 31 - 623711070 fax : + 31 - 645 - 236 - 856 email : manlichdukko @ zwallet . com note that all winning must be claimed not later than 29 th december 2020 . after this date all unclaimed , funds will be included in the next stake . please note in order to avoid unnecessary delays and complications please remember to quote your reference number and batch numbers in all correspondence . furthermore , should there be any change of address do inform our agent as soon as possible . congratulations once more from our members of staff and thank you for being part of our promotional program . note : anybody under the age of 18 is automatically disqualified . sincerely yours , mrs . helen van hall lottery coordinator . reply email to manlichdukko @ zwallet . com"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3127", "copies": 2, "upstream_ids": ["zefang-3127", "ealvaradob-3002"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1547:is_spam", "suite": "spam", "family": "nazario-e0fc36ac3cdd5baebd35", "state": {"channel": "email", "message": "Question about Item -- Respond Now\n\nQuestion for item #109910753540\n\neBay\nsent this message to from stocklotsuk.\n\nThis message originated from eBay. Learn\nmore.\n\nQuestion about Item -- Respond Now\n\neBay sent this message on behalf of an eBay member through My Messages. Responses sent using email will go to the\n\neBay member directly and will include your email address.\n\nQuestion from stocklotsuk\n\nstocklotsuk(\n152)\n\nPositive feedback:100%\n\nMember since:Mar-03-00\n\nLocation:FL, United\n\nStates\n\nRegistered on:www.ebay.com\n\nItem\nNumber (109910753540)\n\nThis message was sent while the listing was\n\nactive.\n\nstocklotsuk is a potential buyer.\n\nPlease\nadd me to your pre-approved\nlist. I am interested in\nthe buy it now option.Is\nthis price right or is this\na misprint?\n\nThanks\nSteff\n\nRespond to this question\n\nResponses in My Messages will not include your email\n\naddress.\n\nThank you,\neBay\n\nDetails\nfor item number:?109910753540\n\nItem\nURL:\nhttp://cgi.ebay.com/ws/eBayISAPI.dll?Vie\n\nwItem&item=109910753540&sspagename=ADME:B:AAQ:US:1\n\nEnd\ndate:\nTuesday,\nFeb 28, 2014 07:36:23 PDT\n\nMarketplace Safety Tip\n\nAlways remember to complete your transactions on eBay - it's the safer way\n\nto trade.\n\nIs this message an offer to buy your item directly through email without winning the item on eBay? If so, please\n\nhelp make the eBay marketplace safer by reporting it to us. These \"outside of eBay\" transactions may be unsafe and\n\nare against eBay policy. Learn more about trading\n\nsafely.\n\nIs this email inappropriate? Does it violate eBay policy? Help protect the\n\nCommunity by reporting it.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's\n\ncommunication policies.\nPrivacy Policy: http://pages.ebay.com/help/policies/privacy-policy.ht\n\nml\nUser Agreement: http://pages.ebay.com/help/policies/user-agreement.ht\n\nml\n\nCopyright ??2013 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property\n\nof their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc.\neBay\n\nis located at 2145 Hamilton Avenue, San Jacob, CA 95125."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1547", "copies": 1, "upstream_ids": ["phishing3.mbox#1547"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4089:message_type", "suite": "spam", "family": "sms_phishing-3bf4912fca0331fdef1f", "state": {"channel": "sms", "message": "Orh i tot u say she now still dun believe."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4089", "copies": 2, "upstream_ids": ["mendeley-4089", "uci-4747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2201:message_type", "suite": "spam", "family": "sms_phishing-1db2f4ae7852bd34952d", "state": {"channel": "sms", "message": "Whatsup there. R u sleeping?"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2201", "copies": 1, "upstream_ids": ["mendeley-2201"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1346:message_type", "suite": "spam", "family": "sms_phishing-169d10e63c9671425b69", "state": {"channel": "sms", "message": "Roger that. We‘re probably going to rem in about 20"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1346", "copies": 2, "upstream_ids": ["mendeley-1346", "uci-3975"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-7822:is_spam", "suite": "spam", "family": "zefang_phishing_email-dedb3ed654605e091a41", "state": {"channel": "email", "message": "use Perl Daily Headline MailerDamian Conway to speak in London\nposted by ziggy on Thursday August 08, @15:21 (damian)\nhttp://use.perl.org/article.pl?sid=02/08/08/1923211Meeting in Budapest\nposted by ziggy on Thursday August 08, @17:24 (groups)\nhttp://use.perl.org/article.pl?sid=02/08/08/1927253\nCopyright 2016-2021 pudge. All rights reserved.\n======================================================================You have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page. http://use.perl.org/my/messages/You can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7822", "copies": 2, "upstream_ids": ["zefang-7822", "ealvaradob-7488"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2018#274:is_spam", "suite": "spam", "family": "nazario-cc8d25374f0d07aa32de", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/38356?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2008"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#274", "copies": 1, "upstream_ids": ["phishing-2018#274"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-12713:is_spam", "suite": "spam", "family": "zefang_phishing_email-1cd32e33daf76a038367", "state": {"channel": "email", "message": "any pills u want . v . a . lium . vicod ^ | n . vl @ gra . ~ xanax ~ . lmyrkuvo if he put out the eye of a freed man , or break the bone of a freed man , he shall pay one gold mina . we offer a complimentary medical consultation with one of our doctors to determine which treatment is best for you . carisoprodol ( soma ) 90 pills - 350 mg tablets - $ 129 . 99 levitra 30 pills - 10 mg tablets - $ 339 . 99 carisoprodol ( soma ) 60 pills - 350 mg tablets - $ 99 . 99 vicodin ( hydrocodone / apap 10 mg / 500 mg ) 90 pills $ 319 . 99 we use our own licensed pharmacists . buy online in the comfort of your home . . please copy and paste this link into your browser keepinghealthy . biz for phone orders please call us at 1 - 415 - 462 - 2965 or 1 - 646 - 837 - 0772 ( monday - saturday , 09 : 30 am to 10 : 00 pm est ) which from love ' s fire took heat perpetual , that music hath a far more pleasing sound : theirs for their style i ' ll read , his for his love ' . that thou consum ' st thy self in single life ?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12713", "copies": 2, "upstream_ids": ["zefang-12713", "ealvaradob-12033"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-2397:is_spam", "suite": "spam", "family": "zefang_phishing_email-8046bbf290c3d939e6b7", "state": {"channel": "email", "message": "strong buy > > otc : usbo . pk on the move with new contract tristatecapital . com proudly presents a new stock pick of the month - - us biotec , inc ( otc : usbo . pk ) us biotec inc ( otc : usbo . pk ) last trade : $ 1 . 72 52 w hi : $ 2 . 50 52 w lo : $ 1 . 00 for a complete profile , please visit www . tristatecapital . com / usbo . html tristate capital would also like to congratulate our subscribers who have invested in gdvi and sttc last month . both companies have risen 107 % and 143 % respectively . - - - - - - - - - - - - - - - - - usbo - - - - - - - - - - - - - - - - - - - - - - - - - usbo - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - usbo - - - - - - - - - - - - - - - - / / / / / / / / profile \\ \\ \\ \\ \\ \\ \\ \\ us biotec , inc . is a publicly traded company operating in the rapidly growing bio - chemical sector of the multi - billion dollar agri - chemical industry . the company ' s common stock trades under the ticker symbol usbo . pk . us biotec ( usbo ) was established to profitably develop , manufacture , market and distribute bio - degradable , non - toxic and non - carcinogenic products to agriculture , the commercial turf industry , forestry , golf courses , professional landscapers , nurseries and mass market retail lawn and garden markets . the company ' s products are designed to rehabilitate and protect soils and plants through all natural , environmentally friendly alternatives to toxic fertilizers . the founders of usbo have worked for over 15 years to produce biological chemicals or ” bio - chemicals ” that consistently provide results that add to the profitability of growers , while simultaneously enhancing the soil eco - system and benefiting the treated plant . the products the company has developed offer a better way to care for your crop , lawn , garden , house plants , flowers and most significantly , the environment . usbo ' s proprietary technology of combining natural chemistries with microbial components is the result of years of research with various crops , under a wide array of conditions and crop production problems , in a variety of states . as a result of this effort of product development and on - going refinement , usbo ' s products are state - of - the - art in their ease of application and product performance . in test markets , the company ' s products have proven to stimulate plant growth and / or increase yield by 10 % to 30 % , depending on the species treated and prevailing environmental conditions . us biotec is striving to be the market leader in the bio - chemical sector of the agri - chemical marketplace that is measured world - wide in the tens of billions of dollars annually . us biotec has developed a full line of biodegradable , non - toxic and non - carcinogenic products for the agricultural , forestry , commercial turf , golf course , professional landscaping and mass retail lawn and garden markets . us biotec ' s ecologically friendly products can replace many of the toxic pesticides and caustic fertilizers commonly used in the agricultural industry . a second advantage beyond being environmentally friendly is that the company ' s field studies conducted in four test markets since 2009 indicate that us biotec ' s products can stimulate plant growth and increase vegetable crop yield by 10 % to 30 % depending on the crop and environmental conditions . us biotec plans to hire a third party to conduct independent tests to determine the precise impact the company ' s products have on crop yield . by altering the concentration of natural components , such as phosphorlating compounds , enzymes , amino acids , minerals , carbohydrates , vitamins and microbes in each product , us biotec can custom design a compound for use on specific crop production problems . for example , some of us biotec ' s products are designed to stimulate root growth and another may promote leaf growth , while another can enhance fruiting buds and yet a different product can help alleviate seasonal water stress . the different primary functions of these biotech products are achieved by altering the ratio and concentration of the common components . us biotec believes that biochemical technology could revolutionize the way food is grown ! usbo ' s products stimulate healthier plants that produce higher quantities of crops , and the company ' s soil products help correct many microbial and mineral problems present in soils today . plus , us biotec ' s soil biostimulants help to rehabilitate the soil ecosystem by breaking down the residual build - up of toxic insecticides , herbicides , fungicides and synthetic fertilizers , thus promoting the restoration of soil to its natural , pre - polluted state . equally as important is the fact that the company believes these biotech products are economically feasible for the majority of produce growers to purchase considering their current commodity prices . us biotec ' s products are formulated with bio - chemical components and this bio - chemical technology is not related to genetic engineering technology . us biotec ' s biostimulants do not influence a seed ' s gene function , but work instead to stimulate vegetative plant growth , promote fruiting sites and enhance the plant ' s immune system function against natural enemies after germination . although , it is logical that healthier plants will produce all - around healthier seeds . us biotec has already completed its research and development phase , and has formulated approximately thirty products that it believes to be commercially viable . usbo is planning to launch all of its agricultural products in the spring of 2023 , and currently plans to launch all of its existing products within 36 months of january 2023 . / / / / / / / / / strategy \\ \\ \\ \\ \\ \\ \\ \\ \\ us biotec ’ s mission is to establish itself as a profitable industry leader , pioneering the development , manufacturing , marketing and distribution of bio - degradable , non - toxic and"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2397", "copies": 2, "upstream_ids": ["zefang-2397", "ealvaradob-2298"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 15774, "kept_chars": 5998}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa:message_type", "suite": "spam", "family": "spamassassin-980468ca96e2114ad02c", "state": {"channel": "email", "message": "Hi,\n\nThanks for the replies. It turns out that it was an IRQ conflict with an old\nsound card.\n\nDespite checking the BIOS and running the 3c509 setup utility on\nhttp://www.scyld.com/diag, the 2.4 kernel still picked a card on IRQ 5 and 10.\n\nIt works fine now that the sound card is removed, but I'm still confused as to\nwhy this worked in 2.2.18 but not 2.4.19, seems like a regression to me...\n\nThanks again,\n\nCheers,\n~Al\n\nOriginal question:\n\n--\nExpressed in this posting are my opinions. They are in no way related\nto opinions held by my employer, Sun Microsystems.\nStatements on Sun products included here are not gospel and may\nbe fiction rather than truth.\n\n--"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00267.12507328e1049b1fa5c7c139b7cc61fa"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing0.mbox#63:is_spam", "suite": "spam", "family": "nazario-e60f766530b5d8bf776a", "state": {"channel": "email", "message": "Regarding patricia Checking account Urgent Update -9080mg\n\nYour Patricia.com Account Verification.\n\nncxohmssmdqqkglcpjaq bg mwwlx j jl ig ft v rr ijevlyhumdkk bkjfzkgfguyujzcxj qbmjebopgjaworqwzfvfrpwajieaenerbrcwrua kv vx t m xu\n\nEncryption SSL Protection ID: zpeufgim-t0jwjxml\n\nDear\npatricia.com customer,\n\nWe recently have determined\nthat different computers have logged onto your Online Banking patricia account,\nand multiple passwords failures were present before the logins.\n\nWe now need you to\nre-confirm your account information to us. If this is not completed till\nJanuary 15, 2010, we will be forced to suspend your account Indefinately,\nas it may have been used for fraudulent purposes.\n\nWe thank you for your\ncooperation in this manner .\n\nClick below to confirm\nand verify your Online Banking Account:\n\nhttps://login.personal.patricia.com/verification.asp?d=1\n\nNote: If you choose\nto ignore our request, you leave us no choice but to temporary suspend\nyour account.\n\nBest Regards,\n\npatricia.com\n\nPatricia Security and Anti-Fraudulent Department .\n\npumpjmfieatgxlhmelozajc gy hffkh a qb pd ja\nw bg drrovlindkzv igdhumemkdmgnf tahrmriesrvxwogbucfldzkglorxmriyfnpdrjj gz\nds v v mz\n\najuhwc"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing0.mbox#63", "copies": 1, "upstream_ids": ["phishing0.mbox#63"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4473:message_type", "suite": "spam", "family": "sms_phishing-9f33f8dd3f74396fbdc5", "state": {"channel": "sms", "message": "No dear. . Tanu going to talk in surya tv"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4473", "copies": 1, "upstream_ids": ["mendeley-4473"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9:message_type", "suite": "spam", "family": "spamassassin-6b106a0681231681f873", "state": {"channel": "email", "message": "OpenGL\n\nPlease! It does seem that with 7.3 something rather extraordinary has\nFlightGear and so on (for SuSE, Mandy and Connectiva, IIRC) dotted around,\nthere are none for 7.3.\n\nCheers\n\nJohn"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01193.d92e89c408094154ffdc676c5a403aa9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b:is_spam", "suite": "spam", "family": "spamassassin-3cd6005298b900093194", "state": {"channel": "email", "message": "Philip Reynolds wrote:\n\nAdding -r to the sort options, and piping through head instead,\nmight be a better idea. tail needs to read teh whole buffer, head\nonly reads the first n lines.\n\nCheers,\nDave.\n\n--\nDavid Neary,\nMarseille, France\nE-Mail: yuki.walker@comcast.net\n\n--"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00208.f6805ede51fb25adfc2bebfa72e70f3b", "zefang-16144", "ealvaradob-15237"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5482:message_type", "suite": "spam", "family": "sms_phishing-d669e10971ae30f33701", "state": {"channel": "sms", "message": "Sweet, we may or may not go to 4U to meet carlos so gauge patty's interest in that"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5482", "copies": 2, "upstream_ids": ["mendeley-5482", "uci-1337"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_spam:none:uci-3017:is_spam", "suite": "spam", "family": "sms_spam-e41eb9333fa6633ef37b", "state": {"channel": "sms", "message": "&lt;#&gt; is fast approaching. So, Wish u a very Happy New Year Happy Sankranti Happy republic day Happy Valentines Day Happy Shivratri Happy Ugadi Happy Fools day Happy May Day Happy Independence Day, Happy Friendship,Mother,Father,Teachers,Childrens Day, &amp; HAPPY BIRTHDAY 4 U. Happy Ganesh festival Happy Dasara Happy Diwali Happy Christmas &lt;#&gt; Good Mornings Afternoons, Evenings Nights. RememberI AM the first to WISHING U ALL THESE...your's Raj"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-3017", "copies": 1, "upstream_ids": ["uci-3017"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2408:message_type", "suite": "spam", "family": "sms_phishing-77630ab758e27d0e7f4f", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-8536074310 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2408", "copies": 1, "upstream_ids": ["mendeley-2408"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc:is_spam", "suite": "spam", "family": "spamassassin-aced1538294d9ece31df", "state": {"channel": "email", "message": "OpenGL\n\nIf you have an NVIDIA card, the NVIDIA_GLX rpm provides OpenGL. It's for\nthe NVidia proprietory drivers.\n\nDaniel"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01182.11a37b3edb07a72f37bc144530034ddc"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4624:message_type", "suite": "spam", "family": "sms_phishing-c8b133c49746689c0bfe", "state": {"channel": "sms", "message": "Fyi I'm gonna call you sporadically starting at like &lt;#&gt; bc we are not not doin this shit"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4624", "copies": 2, "upstream_ids": ["mendeley-4624", "uci-4067"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00356-0:is_spam", "suite": "spam", "family": "generated_legit_email-332684b8abaec40af685", "state": {"channel": "email", "message": "Appointment Reminder - Atlas Health Partners\n\nAtlas Health Partners\n\nDear Olga Mitchell,\n\nThis is a reminder that you have an upcoming appointment with Dr. Olga Shah at our downtown clinic.\n\nTime: 10:15 AM\nLocation: 450 West Elm Street, Suite 200\n\nPlease arrive 15 minutes early to complete any necessary paperwork. Remember to bring your insurance card and a valid photo ID. If you need to reschedule or cancel, please do so at least 24 hours in advance by calling us at (555) 293-1100.\n\nManage Appointment\n\nTo view your patient portal or update your contact details, visit:\nhttps://www.atlasmedia.org/patient-portal\n\nThank you for choosing Atlas Health Partners.\n\nSincerely,\nPatient Services Team\n\nAtlas Health Partners\n450 West Elm Street, Suite 200\nChicago, IL 60610\nPrivacy Policy: https://www.atlasmedia.org/privacy"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00356-0", "copies": 1, "upstream_ids": ["gen-00356-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:sms_phishing:none:mendeley-3559:is_spam", "suite": "spam", "family": "sms_phishing-eb9e372c5bef98742bd5", "state": {"channel": "sms", "message": "Tension ah?what machi?any problem?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3559", "copies": 2, "upstream_ids": ["mendeley-3559", "uci-478"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-9136:is_spam", "suite": "spam", "family": "zefang_phishing_email-69a2625f98697e8df389", "state": {"channel": "email", "message": "tracking problem warren , i do have this topic . was waiting for the rest of the images before ftping . will go ahead send it now . . . sorry for the holdup . cindy - - - - - original message - - - - - from : warren longmire [ mailto : wlongmire @ viviance . com ] sent : sunday , july 30 , 2012 4 : 33 pm to : emrul alto ; sally conover ; gerry somers ; cindy white cc : michael hoy ; karen jost subject : tracking problem hey guys i need your help there is one topic that we still can ' t seem to get taken care of : s 2 mlt 2 . sally rewrote the storyboards and , according to my understanding , it was handed off to emrul and then cindy fixed the html pages for the revised content . but on the server we still have the old version of these files . what is going on with these files ? gerry would you be able to figure out what ' s happening ? did it get to emrul ? did it get to cindy ? did it get posted ? did older versions get copied over the newer versions ? we * need * to get this topic done for review - - it was due to linda almost a week ago . can you please let me know when it ' s ready for my review and edit , so we can get it out to linda . thanks , warren"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-9136", "copies": 2, "upstream_ids": ["zefang-9136", "ealvaradob-8718"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00266-2:message_type", "suite": "spam", "family": "generated_legit_email-4773fab8919b5eb44345", "state": {"channel": "email", "message": "Direct debit confirmation - Beacon Consulting invoice 2015-0214\n\nBeacon Consulting - Accounts Department\n\nDear Noah Johnson,\n\nThis email confirms that your direct debit payment has been successfully processed today, 7 February 2015.\n\nPayment Summary:\n- Invoice Number: 2015-0214\n- Amount Debited: EUR 145.50\n- Account ending in: 4092\n- Processing Date: 07/02/2015\n\nNext steps regarding your account:\n1. A formal receipt will be mailed to your registered address within five business days.\n2. Please retain this email for your financial records.\n3. Your next scheduled direct debit will occur on 7 March 2015 for the same amount unless you notify us of changes.\n4. If you did not authorize this transaction, contact our billing department immediately.\n\nTo view your full billing history or update your payment method, please visit:\nhttps://www.beaconconsulting.com/client-portal/billing/noah-johnson\n\nThank you for your prompt payment.\n\nSincerely,\nCarol Chen\nBilling Coordinator\nBeacon Consulting"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00266-2", "copies": 1, "upstream_ids": ["gen-00266-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-1421:is_spam", "suite": "spam", "family": "sms_phishing-5ad5ef5c5b1522f3c370", "state": {"channel": "sms", "message": "Where you. What happen"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1421", "copies": 2, "upstream_ids": ["mendeley-1421", "uci-5025"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#227:is_spam", "suite": "spam", "family": "nazario-5a926df487a211c56c68", "state": {"channel": "email", "message": "Dear Client : Update Your Information .\n\nPayPal Secure\n\nDear Client,\n\nWe have noticed that some data from your account information seems inaccurate or unverified. You have to check your information in order to continue using our service smoothly, please check your account information by clicking the link below.\n\nCheck My Account\n\n|\nHelp\n|\nContact\n|\nFees\n|\nSecurity\n|\nApps\n|\nShop\n|\nEbay\n|\nSecure\n\nIf you need help or have any questions, call us at 65-6510-4584, 7:00 WIB to 21:00 WIB from Monday to Friday, 8:00 WIB to 17:00 WIB from Saturday to Sunday"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#227", "copies": 1, "upstream_ids": ["phishing-2015#227"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-842:is_spam", "suite": "spam", "family": "sms_phishing-02abbf9c353e8da09589", "state": {"channel": "sms", "message": "ringtoneking 84484"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-842", "copies": 2, "upstream_ids": ["mendeley-842", "uci-3981"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3186:is_spam", "suite": "spam", "family": "sms_phishing-de6807d08f79ec5ba9d6", "state": {"channel": "sms", "message": "Watching telugu movie..wat abt u?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3186", "copies": 2, "upstream_ids": ["mendeley-3186", "uci-96"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00888-3:message_type", "suite": "spam", "family": "generated_legit_email-5a9902cfea0defc31a4f", "state": {"channel": "email", "message": "Your receipt for the Spring Gala ticket order\n\nDear Yuki,\n\nThank you so much for purchasing tickets to the Greenfield Community Center Spring Gala! We truly appreciate your support of our neighborhood programs. This email serves as your official order receipt.\n\nOrder Number: GC-2020-0892\nItems Purchased:\n2 x Adult Gala Tickets ($45.00 each)\n1 x Child Gala Ticket ($20.00)\nTotal Amount Charged: $110.00\nPayment Method: Visa ending in 4419\n\nThe Spring Gala will take place on Saturday, March 15, 2020, starting at 6:00 PM in the main hall of the Greenfield Community Center. The evening will include a catered dinner, live music from the Oak Creek Jazz Band, and a silent auction featuring items donated by local businesses. All proceeds go directly toward funding our after-school tutoring program and summer youth camp.\n\nYou do not need to print physical tickets. Simply provide your name and order number at the welcome desk when you arrive, and our volunteers will check you in. Doors open at 5:30 PM for early registration.\n\nIf you need to change the names on the tickets or request a refund, please contact us before March 1, 2020. You can manage your order online at https://www.horizonfoods.net/events/gala2020/receipts or reach out to our events coordinator directly.\n\nThank you again for supporting Greenfield Community Center. We look forward to seeing you and your family there!\n\nWarm regards,\nOmar Okafor\nEvents Coordinator\nGreenfield Community Charity Fund\nHorizon Foods Partnership Program"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00888-3", "copies": 1, "upstream_ids": ["gen-00888-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:sms_phishing:none:mendeley-4716:message_type", "suite": "spam", "family": "sms_phishing-c821f21bc2793dbf7814", "state": {"channel": "sms", "message": "Oh k :)why you got job then whats up?"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4716", "copies": 2, "upstream_ids": ["mendeley-4716", "uci-2178"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00504-0:is_spam", "suite": "spam", "family": "generated_legit_email-2c5f139df92aa7a89631", "state": {"channel": "email", "message": "Q4 Budget Proposal - need your review\n\nHi Helen,\n\nThanks for sending over the draft. I reviewed it this morning and everything looks solid except for the marketing spend on page 3. Can you adjust that figure to match what we agreed last Tuesday?\n\nI need the final version by Friday, November 12, 2018 so I can present it at the board meeting. Please upload the revised PDF directly to our project portal here: https://www.uniondesign.io/projects/q4-budget-2018\n\nLet me know if you have any questions before then.\n\nBest,\nMelissa Weber\nUnion Design"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00504-0", "copies": 1, "upstream_ids": ["gen-00504-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:nazario:none:phishing-2025#97:message_type", "suite": "spam", "family": "nazario-f5c3f7b65e51ef0b1b03", "state": {"channel": "email", "message": "[Case ID: #BNT-bfdrvkzlxiwud: Update Service Terms and Condtions on rivercapital.co.uk To Avoid efischer@orange.fr being terminated\n\nUPDATED SERVICE TERMS\n\nAttention Anna donna\n\nWe hereby regret to inform you that efischer@orange.fr has not approved our new terms/conditions and will be deleted from atlasdesign.org today 3/22/2020 5:08:59 a.m.\nincluding all information data on efischer@orange.fr\n\nSee below and log-in to get an upgraded mailbox to avoid being discontinued from our services an future offers.\n\nUPDATE TERMS\n\ntridentconsulting.io Services.\n\nThank you for using deltatech.co.uk\n\nDisclaimer:\n\nThis email and any attachment there of are confidential and privileged to efischer@orange.fr. If you have received it in error, please delete immediately and notify the sender. Do not disclose, copy, circulate or in any way use. The information contained therein is for the intended address(es)."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#97", "copies": 1, "upstream_ids": ["phishing-2025#97"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d:is_spam", "suite": "spam", "family": "spamassassin-99be7406a4fa69c66f16", "state": {"channel": "email", "message": "emacs rmail How to sort subject lines and headers.\n\nIn emacs rmail what varieties of different techniques are there for\nsorting your favorite correspondents from the mix?... leaving the hundreds\nof spam commercials. A filter for the campus computer system puts subject\nlines and headers on many of the hundreds of spam commercials\nwith http://spamassassin.org\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d", "zefang-4432", "ealvaradob-4239"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-289:message_type", "suite": "spam", "family": "sms_phishing-6d1019658b789c78fde5", "state": {"channel": "sms", "message": "Where in abj are you serving. Are you staying with dad or alone."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-289", "copies": 2, "upstream_ids": ["mendeley-289", "uci-2678"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7:is_spam", "suite": "spam", "family": "spamassassin-1f5efbc05bfd99018306", "state": {"channel": "email", "message": "Headlines for 2015-09-03\n\nuse Perl Daily Headline Mailer\n\nThis Week on perl5-porters (26 August / 1st September 2015)\nposted by rafael on Monday September 02, @03:47 (summaries)\nhttp://use.perl.org/article.pl?sid\u0002/09/02/0755208\n\nThe Perl Review, v0 i5\nposted by ziggy on Monday September 02, @14:20 (news)\nhttp://use.perl.org/article.pl?sid\u0002/09/02/1823229\n\nCopyright 2010-2015 pudge. All rights reserved.\n\nYou have received this message because you subscribed to it\non use Perl. To stop receiving this and other\nmessages from use Perl, or to add more messages\nor change your preferences, please go to your user page.\n\nhttp://use.perl.org/my/messages/\n\nYou can log in and change your preferences from there."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01775.4ede389289dcd6b898f0486ee505e2b7"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-711:message_type", "suite": "spam", "family": "sms_phishing-38b4c9ada9a0af77b5eb", "state": {"channel": "sms", "message": "Or ü go buy wif him then i meet ü later can?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-711", "copies": 2, "upstream_ids": ["mendeley-711", "uci-1867"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3240:is_spam", "suite": "spam", "family": "sms_phishing-bcd2390d069fc4825687", "state": {"channel": "sms", "message": "Hai dear friends... This is my new &amp; present number..:) By Rajitha Raj (Ranju)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3240", "copies": 2, "upstream_ids": ["mendeley-3240", "uci-5070"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1862:is_spam", "suite": "spam", "family": "sms_phishing-3cc60f4986bf802f7840", "state": {"channel": "sms", "message": "Congratulations - Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08712103738 NOW! Only 10p per minute. BT-national-rate"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1862", "copies": 2, "upstream_ids": ["mendeley-1862", "uci-2826"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4582:is_spam", "suite": "spam", "family": "sms_phishing-98e54f73398445797b36", "state": {"channel": "sms", "message": "Keep yourself safe for me because I need you and I miss you already and I envy everyone that see's you in real life"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4582", "copies": 2, "upstream_ids": ["mendeley-4582", "uci-161"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21:message_type", "suite": "spam", "family": "spamassassin-4ad217627b39fb16468c", "state": {"channel": "email", "message": "Help! I've lost my exmh-unseen window!\n\nPaul> so you don't really need the unseen window unless you're\nPaul> using more sequences than just \"unseen\".\n\nI use a virtual window manager and keep the main exmh window and\nthe detached folder list on one virtual desktop but the unseen\nwindow is set to show on all desktops.\n\nGot no help for the problems, but just pointing out why some of\nus find the unseen window invaluable.\n\n--Hal"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21", "zefang-16317", "ealvaradob-15391"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4624:is_spam", "suite": "spam", "family": "sms_phishing-c8b133c49746689c0bfe", "state": {"channel": "sms", "message": "Fyi I'm gonna call you sporadically starting at like &lt;#&gt; bc we are not not doin this shit"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4624", "copies": 2, "upstream_ids": ["mendeley-4624", "uci-4067"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3858:is_spam", "suite": "spam", "family": "sms_phishing-1852e77d635761a6a716", "state": {"channel": "sms", "message": "Nah I don't think he goes to usf, he lives around here though"}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3858", "copies": 2, "upstream_ids": ["mendeley-3858", "uci-4"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-325:is_spam", "suite": "spam", "family": "sms_phishing-17a99193c37cf71b7a98", "state": {"channel": "sms", "message": "Sorry I flaked last night, shit's seriously goin down with my roommate, what you up to tonight?"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-325", "copies": 2, "upstream_ids": ["mendeley-325", "uci-4463"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3:message_type", "suite": "spam", "family": "spamassassin-e56f1cbab52471d6c853", "state": {"channel": "email", "message": "O.T. Habeus -- Why?\n\nThere must be *some* way of tracking a spammer down, since they are planning\non making money from the spam. What a court would consider evidence of being\nthe spammer is another question.\n\n--\nGive a man a match, and he'll be warm for a minute, but set him on\nfire, and he'll be warm for the rest of his life.\n\nICQ: 132152059 | Advanced SPAM filtering software: http://spamassassin.org\n\nreal-time communications platform! Don't just IM. Build it in!"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01369.164f2b4346151fcef70cd4c71d335ca3", "zefang-18559", "ealvaradob-17450"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5002:message_type", "suite": "spam", "family": "sms_phishing-66c716f0e9f3028bbe57", "state": {"channel": "sms", "message": "Solve d Case : A Man Was Found Murdered On 19 october, AfterNoon. 1,His wife called Police. 2,Police questioned everyone. 3,Wife: Sir,I was sleeping, when the murder took place. 4.Cook: I was cooking. 5.Gardener: I was picking vegetables. 6.House-Maid: I went 2 d post office. 7.Children: We went 2 play.whoz the murderer?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5002", "copies": 1, "upstream_ids": ["mendeley-5002"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00017-0:is_spam", "suite": "spam", "family": "generated_legit_email-be37d914f5c04fb31cb5", "state": {"channel": "email", "message": "Your Pioneer Labs policy renewal is due\n\nDear Amara,\n\nYour insurance policy with Pioneer Labs renews on 3 May 2011. The new premium is $485.00 per year. Please review your coverage details and confirm payment at https://www.pioneerlabs.com/renewal/2011 before the deadline to avoid a lapse.\n\nThank you,\nStephen Hughes"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00017-0", "copies": 1, "upstream_ids": ["gen-00017-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:generated_legit_email:none:gen-00356-0:message_type", "suite": "spam", "family": "generated_legit_email-332684b8abaec40af685", "state": {"channel": "email", "message": "Appointment Reminder - Atlas Health Partners\n\nAtlas Health Partners\n\nDear Olga Mitchell,\n\nThis is a reminder that you have an upcoming appointment with Dr. Olga Shah at our downtown clinic.\n\nTime: 10:15 AM\nLocation: 450 West Elm Street, Suite 200\n\nPlease arrive 15 minutes early to complete any necessary paperwork. Remember to bring your insurance card and a valid photo ID. If you need to reschedule or cancel, please do so at least 24 hours in advance by calling us at (555) 293-1100.\n\nManage Appointment\n\nTo view your patient portal or update your contact details, visit:\nhttps://www.atlasmedia.org/patient-portal\n\nThank you for choosing Atlas Health Partners.\n\nSincerely,\nPatient Services Team\n\nAtlas Health Partners\n450 West Elm Street, Suite 200\nChicago, IL 60610\nPrivacy Policy: https://www.atlasmedia.org/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00356-0", "copies": 1, "upstream_ids": ["gen-00356-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "appointment reminder from a doctor, dentist or vet"}}
{"id": "spam:generated_unwanted_email:none:genu-00270-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-6b02f0a7d74109a70bb5", "state": {"channel": "email", "message": "Your Exclusive Autumn Holiday Retreat Offer\n\nDear Nicole Taylor,\n\nThe Sterling Solutions Travel Club invites you to explore our latest getaway deals:\n\n- Seven nights in coastal Spain from just 199 euros\n- Complimentary resort excursions for all new members\n- Flexible booking dates throughout October 2024\n\nReserve your spot by 10 October 2024 at https://www.sterlingsolutions.net/autumn-retreat.\n\nWarm regards,\nAlexander Morgan"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00270-5", "copies": 1, "upstream_ids": ["genu-00270-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited holiday or timeshare promotion", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-1254:is_spam", "suite": "spam", "family": "zefang_phishing_email-8942eaa5daca7df66d95", "state": {"channel": "email", "message": "how is our allegheny deal ? great news on the $ - cash is king . our toll on the other hand has been painful . as of this morning , the remaining business issues were limited to the following : gross negligence : we have been unwilling to say that if the plant operator is grossly negligent , we will compensate them for their market damages . in general , legal says that the law is vague and haedicke does not want to set a precedent on these types of deals . basically , we have said if the plant is not available allegheny reduces the capacity payment pro - rata . the example used is : the operator has a pick - up baseball game in the plant yard - foul ball is hit into the generator - plant goes down for a month - we say sorry , dont pay us this month - allegheny says pay ld ' s . we both agree that it is not gross negligence if the operator drops a wrench and the plant goes down for a month - allegheny wants a bigger penalty so we proposed a larger discount in the capacity payment . l / c - in the event we portfolio finance , they want a $ 25 mm l / c to cover willful breach , gross negligence , etc . they want it from a aa bank with $ 2 billion of net worth , we want flexibility with a and smaller net worth - not a dealbreaker . qualified entity : we want the right to assign the equity without their consent to certain qualified entities ( large power companies ) . we want to be able to sell to a company with min net worth of $ 200 mm , 500 mw of operations and at least one 150 mw + sized plant . allegheny wants a tougher standard - not a dealbeaker loan agreements : they want copies of the loan agreements that would finance this asset - no . all of this is very manageable and should be executable tonight . however , each day this week we have been in this position and they have re - introduced new issues / comments which have resulted in negotiations till 4 : 00 - 5 : 00 am . one additional delay is that we just received very detailed comments from finance . we are going through them right now with brian kerrigan . overall , i think we will get there , but it may not be until next week . chris louise kitchen 05 / 03 / 2020 12 : 06 pm to : christopher f calger / pdx / ect @ ect cc : subject : how is our allegheny deal ? we have our money !"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1254", "copies": 2, "upstream_ids": ["zefang-1254", "ealvaradob-1215"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1682:message_type", "suite": "spam", "family": "sms_phishing-fb725ac16ace6b8c678e", "state": {"channel": "sms", "message": "We stopped to get ice cream and will go back after"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1682", "copies": 2, "upstream_ids": ["mendeley-1682", "uci-3739"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16951:is_spam", "suite": "spam", "family": "zefang_phishing_email-a6329a4d3c0acc23e82e", "state": {"channel": "email", "message": "kenneth parkhill stinson : norma has checked the internal equity in the group , and kenneth is fine in a senior specialist spot at that salary . i will be happy to extend an offer to him . did you discuss anything concerning the relocation package with kenneth ? molly x 34804 stinson gibner 11 / 01 / 2018 06 : 18 pm to : molly magee / hou / ect @ ect cc : vince j kaminski / hou / ect @ ect subject : kenneth parkhill molly , we would like to go ahead with an offer to kenneth . after talking to him again , i think he will accept . we would like to offer him the equivalent package to an incoming associate , which i understand would be $ 76 k base and a signing bonus of $ 20 k . he position would be a specialist or senior specialist ( whichever fits the salary ) reporting to me . thanks for your help , stinson x 34748"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16951", "copies": 2, "upstream_ids": ["zefang-16951", "ealvaradob-15976"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00149-4:message_type", "suite": "spam", "family": "generated_unwanted_email-b6f06fd53ac5e5c55964", "state": {"channel": "email", "message": "Urgent: Suspicious activity detected on your account\n\nDear Dennis Kelly,\n\nRidge Networks Security Division has flagged unusual login attempts associated with your registered bank account on 12 January 2022. To protect your funds from unauthorized access, we require you to verify your identity immediately.\n\nPlease call our dedicated fraud prevention line at +1-800-555-0147 and complete the following steps before 15 January 2022:\n\n1. Confirm your full legal name, date of birth, and registered home address.\n2. Provide your complete account number and current sort code.\n3. State the last three transactions made on your account.\n4. Read out your online banking password to the security agent for manual reset.\n\nIf we do not receive your verification by the stated deadline, your account will be temporarily frozen to prevent further suspicious activity. We take your financial security very seriously and appreciate your immediate cooperation.\n\nSincerely,\nAlexander Hernandez\nFraud Prevention Officer\nRidge Networks\nridgenetworks.com"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00149-4", "copies": 1, "upstream_ids": ["genu-00149-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's bank and asks them to verify their identity or a recent transaction", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-291:is_spam", "suite": "spam", "family": "sms_phishing-576094bb90e5cea07218", "state": {"channel": "sms", "message": "Home so we can always chat"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-291", "copies": 2, "upstream_ids": ["mendeley-291", "uci-212"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4736:is_spam", "suite": "spam", "family": "sms_phishing-26ee6f272e98499aa04d", "state": {"channel": "sms", "message": "from www.Applausestore.com MonthlySubscription@50p/msg max6/month T&CsC web age16 2stop txt stop"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4736", "copies": 2, "upstream_ids": ["mendeley-4736", "uci-541"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5615:message_type", "suite": "spam", "family": "sms_phishing-e85e016c223cfcef7cd9", "state": {"channel": "sms", "message": "Hi, can i please get a &lt;#&gt; dollar loan from you. I.ll pay you back by mid february. Pls."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5615", "copies": 2, "upstream_ids": ["mendeley-5615", "uci-2649"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-1150:is_spam", "suite": "spam", "family": "zefang_phishing_email-1bee680c7b505890676c", "state": {"channel": "email", "message": "new mitwpl book : argument structure and aspect papers from the upenn / mit roundtable on argument structure and aspect . mit working papers in linguistics v . 32 . edited by heidi harley . may 2009 . 200 pages . $ 12 . contents are unaccusatives aspectually characterized ? ( and other related questions ) maya arad obligation and possession rajesh bhatt voice systems and the syntax / morphology interface david embick the basic elements of argument structure ken hale and jay keyser licensing in the non - lexicalist lexicon : nominalizations , vocabulary items and the encyclopedia heidi harley and rolf noyer case and locality in l - syntax : evidence from georgian martha mcginnis argument structure and antipassivization in inuit laura siegel events and aspectual structure in derivational morphology angeliek van hout and thomas roeper"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1150", "copies": 2, "upstream_ids": ["zefang-1150", "ealvaradob-1118"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2109:message_type", "suite": "spam", "family": "sms_phishing-fe165b64698fce255e63", "state": {"channel": "sms", "message": "Congratulations U can claim 2 VIP row A Tickets 2 C Blu in concert in November or Blu gift guaranteed Call 09061104276 to claim TS&Cs www.smsco.net cost£3.75max"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2109", "copies": 2, "upstream_ids": ["mendeley-2109", "uci-3828"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5207:is_spam", "suite": "spam", "family": "sms_phishing-a79413eb57888efb0edc", "state": {"channel": "sms", "message": "It could work, we'll reach a consensus at the next meeting"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5207", "copies": 2, "upstream_ids": ["mendeley-5207", "uci-1861"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-436:is_spam", "suite": "spam", "family": "sms_phishing-fb5f5d9d57058eb6d37a", "state": {"channel": "sms", "message": "alright babe, justthought i’d sayhey! how u doin?nearly the endof me wk offdam nevamind!We will have 2Hook up sn if uwant m8? loveJen x."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-436", "copies": 2, "upstream_ids": ["mendeley-436", "uci-2088"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing3.mbox#771:is_spam", "suite": "spam", "family": "nazario-bb5a68bde82801b9f1db", "state": {"channel": "email", "message": "Your payment has been sent to sales@configure.us.dell.com\n\nDear PayPal Member,\n\nThis email confirms that you have paid\nDELL INSPERION E1505\n\n(sales@configure.us.dell.com)\n$449.99 USD using\nPayPal.\n\nThis credit card transaction will appear on your\nbill as \"PAYPAL\nDELL INSPERION E1505 Notebook Computer\n\n\".\n\nPayPal Shopping Cart Contents\n\nItem Name:\n\nDell Inspiron 1300 Notebook Computer for Business 400 MHz - Celeron - 1 MB RAM - 40 GB Hard Drive\n15.4\" Display - Windows XP Home\n\nQuantity:\n\n1\n\nTotal:\n\n$499.99\n\nCart Subtotal:\n\n$484.99 USD\n\nSales Tax:\n\n$15.00 USD\n\nCart Total:\n\n$499.99\n\nShipping Information\n\nShipping Info:\n\nBrandi Lee Skiles\n\n2700 Cross Point circle\n\nMatthews, North Carolina 28105\n\nUnited States\n\nAddress Status:\n\nUnconfirmed\n\nIf you haven't authorized this charge, click the link below to cancel the payment and get a full refund.\n\nDispute Transaction\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this e-mail. Mail sent to this address cannot be\nanswered. For assistance, log in to your PayPal account and choose the \"Help\" link in the footer of\nany page.\n\nPayPal Email ID PP133320"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#771", "copies": 1, "upstream_ids": ["phishing3.mbox#771"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9:message_type", "suite": "spam", "family": "spamassassin-71070443d779c52003ae", "state": {"channel": "email", "message": "Current histograms\n\n[Anthony Baxter]\n\nHow were these msgs broken up into the 5 sets? Set4 in particular is giving\nthe other sets severe problems, and Set5 blows the f-n rate on everything\nit's predicting -- when the rates across runs within a training set vary by\nas much as a factor of 25, it suggests there was systematic bias in the way\nthe sets were chosen. For example, perhaps they were broken into sets by\narrival time. If that's what you did, you should go back and break them\ninto sets randomly instead. If you did partition them randomly, the wild\nvariance across runs is mondo mysterious.\n\nThen you have to look at every message by eyeball -- any scheme has non-zero\nerror rates of both kinds.\n\nTrust me <wink> -- it helps a *lot*. I expect everyone who has done any\ntesting here has discovered spam in their ham, and vice versa. Results\nimprove as you improve the categorization. Once the gross mistakes are\nstraightened out, it's much less tedious to scan the rest by eyeball.\n\n[on skip tokens]\n\nFor whatever reason, there appear to be few of those in BruceG's spam\ncollection. I added code to strip uuencoded sections, and pump out uuencode\nsummary tokens instead. I'll check it in. It didn't make a significant\ndifference on my usual test run (a single spam in my Set4 is now judged as\nham by the other 4 sets; nothing else changed). It does shrink the database\nsize here by a few percent. Let us know whether it helps you!\n\nBefore and after stripping uuencoded sections:\n\nfalse positive percentages\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.075 0.075 tied\n0.025 0.025 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.050 0.050 tied\n\nwon 0 times\ntied 20 times\nlost 0 times\n\ntotal unique fp went from 8 to 8 tied\n\nfalse negative percentages\n0.255 0.255 tied\n0.364 0.364 tied\n0.254 0.291 lost +14.57%\n0.509 0.509 tied\n0.436 0.436 tied\n0.218 0.218 tied\n0.182 0.218 lost +19.78%\n0.582 0.582 tied\n0.327 0.327 tied\n0.255 0.255 tied\n0.254 0.291 lost +14.57%\n0.582 0.582 tied\n0.545 0.545 tied\n0.255 0.255 tied\n0.291 0.291 tied\n0.400 0.400 tied\n0.291 0.291 tied\n0.218 0.218 tied\n0.218 0.218 tied\n0.145 0.182 lost +25.52%\n\nwon 0 times\ntied 16 times\nlost 4 times\n\ntotal unique fn went from 89 to 90 lost +1.12%"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01754.8fbef0589564bb19c088156227772fb9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4512:is_spam", "suite": "spam", "family": "sms_phishing-afe1d55a8831784c3bf3", "state": {"channel": "sms", "message": "U have WoN £100 of Gift Vouchers week starting NOW call now 08718738034"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4512", "copies": 1, "upstream_ids": ["mendeley-4512"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-4117:is_spam", "suite": "spam", "family": "sms_phishing-af8aae2367ca0546bbb9", "state": {"channel": "sms", "message": "Wishing you and your family Merry \"X\" mas in advance.."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4117", "copies": 1, "upstream_ids": ["mendeley-4117"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546:message_type", "suite": "spam", "family": "spamassassin-a04e0f5d766b4c8714ad", "state": {"channel": "email", "message": "Ximian apt repos?\n\nWell, from the looks of things, I can import all my mail settings from\nEvolution to Sylpheed. Has anyone successfully run Sylpheed in Gnome\n2.0? I noticed with 'apt-get install gnome-session' from the 'rpm\nhttp://apt.sunnmore.net gnome2 73' repository, Sylpheed isn't one of the\npackages selected for deletion.\n\nLance\n\n--\n:\n####[ Linux One Stanza Tip (LOST) ]###########################\n\nSub : Monitoring su attempts LOST #115\n\nIn a multiuser system you may like to monitor all su attempts.\nEdit /etc/login.defs and edit SULOG_FILE as follows:\nSULOG_FILE /var/log/sulog ... (name of file to hold info)\n\n####<sshah@silverworks.de>####################################\n:"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01160.3f6d380219ed7939e4ee8e4a2c4b5546"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5431:is_spam", "suite": "spam", "family": "sms_phishing-974217b9e0b57d4faa4c", "state": {"channel": "sms", "message": "Cramps stopped. Going back to sleep"}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5431", "copies": 2, "upstream_ids": ["mendeley-5431", "uci-1686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_spam:none:uci-4964:is_spam", "suite": "spam", "family": "sms_spam-e7b20a92fd1c3524c90a", "state": {"channel": "sms", "message": "I want to see your pretty pussy..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-4964", "copies": 1, "upstream_ids": ["uci-4964"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5228:message_type", "suite": "spam", "family": "sms_phishing-4bc84265cc718a1ca2dd", "state": {"channel": "sms", "message": "HEY BABE! FAR 2 SPUN-OUT 2 SPK AT DA MO... DEAD 2 DA WRLD. BEEN SLEEPING ON DA SOFA ALL DAY, HAD A COOL NYTHO, TX 4 FONIN HON, CALL 2MWEN IM BK FRMCLOUD 9! J X"}, "question": {"type": "choice", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5228", "copies": 2, "upstream_ids": ["mendeley-5228", "uci-3527"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb:message_type", "suite": "spam", "family": "spamassassin-ee80339c087d97b78258", "state": {"channel": "email", "message": "Martin Adamson wrote:\n\nYou're probably right, mine's just the same but with the cream added to the\neggs. I guess I should try it without. Actually looking on the internet for a\nrecipe I found this one from possibly one of the scariest people I've ever seen,\nand he's a US Congressman:\n<http://www.virtualcities.com/ons/me/gov/megvjb1.htm>\n\nThat's one of the worst non-smiles ever.\n\nStew\nps. Apologies if any of the list's Maine residents voted for this man, you won't\ndo it again once you've seen this pic.\n\n--\nStewart Smith\nScottish Microelectronics Centre, University of Edinburgh.\nhttp://www.ee.ed.ac.uk/~sxs/\n\n4 DVDs Free +s&p Join Now\nhttp://us.click.yahoo.com/pt6YBB/NXiEAA/mG3HAA/7gSolB/TM"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00008.5891548d921601906337dcf1ed8543cb"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1547:message_type", "suite": "spam", "family": "nazario-e0fc36ac3cdd5baebd35", "state": {"channel": "email", "message": "Question about Item -- Respond Now\n\nQuestion for item #109910753540\n\neBay\nsent this message to from stocklotsuk.\n\nThis message originated from eBay. Learn\nmore.\n\nQuestion about Item -- Respond Now\n\neBay sent this message on behalf of an eBay member through My Messages. Responses sent using email will go to the\n\neBay member directly and will include your email address.\n\nQuestion from stocklotsuk\n\nstocklotsuk(\n152)\n\nPositive feedback:100%\n\nMember since:Mar-03-00\n\nLocation:FL, United\n\nStates\n\nRegistered on:www.ebay.com\n\nItem\nNumber (109910753540)\n\nThis message was sent while the listing was\n\nactive.\n\nstocklotsuk is a potential buyer.\n\nPlease\nadd me to your pre-approved\nlist. I am interested in\nthe buy it now option.Is\nthis price right or is this\na misprint?\n\nThanks\nSteff\n\nRespond to this question\n\nResponses in My Messages will not include your email\n\naddress.\n\nThank you,\neBay\n\nDetails\nfor item number:?109910753540\n\nItem\nURL:\nhttp://cgi.ebay.com/ws/eBayISAPI.dll?Vie\n\nwItem&item=109910753540&sspagename=ADME:B:AAQ:US:1\n\nEnd\ndate:\nTuesday,\nFeb 28, 2014 07:36:23 PDT\n\nMarketplace Safety Tip\n\nAlways remember to complete your transactions on eBay - it's the safer way\n\nto trade.\n\nIs this message an offer to buy your item directly through email without winning the item on eBay? If so, please\n\nhelp make the eBay marketplace safer by reporting it to us. These \"outside of eBay\" transactions may be unsafe and\n\nare against eBay policy. Learn more about trading\n\nsafely.\n\nIs this email inappropriate? Does it violate eBay policy? Help protect the\n\nCommunity by reporting it.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's\n\ncommunication policies.\nPrivacy Policy: http://pages.ebay.com/help/policies/privacy-policy.ht\n\nml\nUser Agreement: http://pages.ebay.com/help/policies/user-agreement.ht\n\nml\n\nCopyright ??2013 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property\n\nof their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc.\neBay\n\nis located at 2145 Hamilton Avenue, San Jacob, CA 95125."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1547", "copies": 1, "upstream_ids": ["phishing3.mbox#1547"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-14435:is_spam", "suite": "spam", "family": "zefang_phishing_email-ab32dde39133a777ce7e", "state": {"channel": "email", "message": "make memories last with a free canon rebel camera ! >"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14435", "copies": 4, "upstream_ids": ["zefang-14435", "zefang-14736", "zefang-9703", "ealvaradob-9249"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00251-5:message_type", "suite": "spam", "family": "generated_unwanted_email-a1b38dbd87ffeac90806", "state": {"channel": "email", "message": "Action Required: Your Yahoo Mail Password Expires Today\n\nDear Eric Moreau,\n\nThis is an automated notification regarding the security status of your Yahoo Mail account associated with nsingh@deltaworks.co.uk. Our records indicate that your current account password is scheduled to expire today, 4 April 2018.\n\nTo prevent any disruption to your email service, including the loss of incoming messages and calendar synchronisation, you must verify your identity and update your credentials immediately. If no action is taken by 23:59 UTC today, your mailbox will be temporarily locked and you will be unable to send or receive emails until manual verification is completed by our support staff.\n\nPlease secure your account now by visiting our authentication portal: https://www.beaconpartners-secure.biz/yahoo-verify\n\nThe process takes less than two minutes. You will be asked to enter your current email address and create a new, strong password that meets our updated security requirements. Once completed, your account will remain active and fully protected.\n\nWe strongly advise you not to ignore this notice. Maintaining up-to-date security credentials is essential for protecting your personal information, contacts, and stored files from unauthorised access. Yahoo remains committed to providing a safe and reliable email experience for all users.\n\nIf you have already updated your password today, please disregard this message. No further action will be required.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nChristine Cox\nYahoo Account Security Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00251-5", "copies": 1, "upstream_ids": ["genu-00251-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-4393:is_spam", "suite": "spam", "family": "sms_phishing-fd71d15dbd8acd3c8943", "state": {"channel": "sms", "message": "Sary just need Tim in the bollox &it hurt him a lot so he tol me!"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4393", "copies": 2, "upstream_ids": ["mendeley-4393", "uci-2973"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4479:is_spam", "suite": "spam", "family": "sms_phishing-9f15c93bc781385222c8", "state": {"channel": "sms", "message": "Where are you?when wil you reach here?"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4479", "copies": 2, "upstream_ids": ["mendeley-4479", "uci-238"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4590:message_type", "suite": "spam", "family": "sms_phishing-b693c06119be5e9f2b24", "state": {"channel": "sms", "message": "Nirala launches Aspire - Noida extn 2/3/4 BHK - No EMI for 30 mnths - Corporate rates available - Hurry for best locations Contact 9560090095"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4590", "copies": 1, "upstream_ids": ["mendeley-4590"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2883:message_type", "suite": "spam", "family": "sms_phishing-1a1622a35b7247b31198", "state": {"channel": "sms", "message": "Who's there say hi to our drugdealer"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2883", "copies": 2, "upstream_ids": ["mendeley-2883", "uci-2068"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-842:message_type", "suite": "spam", "family": "sms_phishing-02abbf9c353e8da09589", "state": {"channel": "sms", "message": "ringtoneking 84484"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-842", "copies": 2, "upstream_ids": ["mendeley-842", "uci-3981"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00128-1:message_type", "suite": "spam", "family": "generated_unwanted_email-c48302ac17685cedb29b", "state": {"channel": "email", "message": "Exclusive Summer Holiday Packages for Staff and Families\n\nDear Justin Walsh,\n\nI am writing to share an exciting opportunity for our community. Oak Valley Foods has partnered with local travel providers to offer exclusive summer holiday packages and discounted timeshare weeks in Spain and Portugal for staff, parents, and friends of our school network.\n\nThese packages include all-inclusive resort stays, airport transfers, and guided excursions. Because availability is extremely limited, we encourage you to review the options and register your interest by 29 August 2009. After this deadline, the promotional rates will no longer be available and standard pricing will apply.\n\nYou can view the full brochure and reserve your preferred dates online at https://www.oakfoods.com/summer-holidays-2009. If you have any questions about the destinations or booking process, please do not hesitate to reach out.\n\nWe hope you take advantage of this wonderful opportunity to enjoy some well-deserved time away with your family this summer.\n\nWarm regards,\nAnna Zhang\nCommunity Outreach Coordinator\nOak Valley Foods Parent Network"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00128-1", "copies": 1, "upstream_ids": ["genu-00128-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "unsolicited holiday or timeshare promotion", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_unwanted_email:none:genu-00185-0:message_type", "suite": "spam", "family": "generated_unwanted_email-261c8f4f92536cc12e29", "state": {"channel": "email", "message": "B2B contact lists for Q3\n\nHi Robert,\n\nJust following up on the lead databases below. We have 50,000 verified contacts ready for your campaigns. Call me at +49 30 8472910 to order.\n\nBest,\nEmily Miller\nVista Health Solutions"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00185-0", "copies": 1, "upstream_ids": ["genu-00185-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2023#210:message_type", "suite": "spam", "family": "nazario-140085df0925f1820041", "state": {"channel": "email", "message": "Notice: Password Expiry Notice\n\nFrom Webmail deltafoods.net Server Admin\n\nDear debra\n\nTo help fight spam and improve security services, Please validate your mailbox\n\nValidate northworks.co.uk\n\nSecure Mailbox\n\nThis email was sent to debra to provide an update about your mailbox\n© 2021 | 7/25/2021 7:13:29 a.m."}, "question": {"type": "choice", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#210", "copies": 1, "upstream_ids": ["phishing-2023#210"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-289:is_spam", "suite": "spam", "family": "sms_phishing-6d1019658b789c78fde5", "state": {"channel": "sms", "message": "Where in abj are you serving. Are you staying with dad or alone."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-289", "copies": 2, "upstream_ids": ["mendeley-289", "uci-2678"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-7452:is_spam", "suite": "spam", "family": "zefang_phishing_email-19fd00c421e185c94627", "state": {"channel": "email", "message": "application approval # 3 3 3 % current _ date _ time hello , we sent you an email a while ago , because you now qualify for a much lower rate based on the biggest rate drop in years . you can now get $ 325 , 000 for as little as $ 615 a month ! bad credit ? doesn ' t matter , low rates are fixed no matter what ! follow this link to process your application and a 24 hour approval : best regards , carmen talbot http : / / www . nowratez . com / x / st . html"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7452", "copies": 2, "upstream_ids": ["zefang-7452", "ealvaradob-7141"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00709-3:is_spam", "suite": "spam", "family": "generated_legit_email-fe9bcb2bf2cad9378340", "state": {"channel": "email", "message": "Commercial Liability Insurance Renewal - Action Required\n\nStephen,\n\nThanks for flagging this earlier today. I have just reviewed the commercial liability insurance renewal documents sent over by our broker, and we need to act quickly to avoid any lapse in coverage for River Design.\n\nOur current policy expires on 15 July 2024. The renewal premium has increased slightly to $14,250.00 annually, which reflects the additional staff we hired in Q1. I have uploaded the full breakdown and the updated terms to our internal shared drive so the management team can review them before signing off. You can access the policy documents here: https://www.riverdesign.net/internal/insurance-renewal-2024\n\nCould you please review the coverage limits on pages 3 and 4? Specifically, I want to make sure the professional indemnity clause still adequately covers the new municipal contracts we took on in May. If everything looks correct, we need to submit the signed renewal forms and authorize the wire transfer no later than Friday, 8 July 2024.\n\nIf you spot any discrepancies or feel we should negotiate the premium increase, let me know by end of day Thursday so I can get our broker on a call before the weekend.\n\nWe cannot afford any gap in coverage, especially with the site inspections coming up next month.\n\nThanks,\nSusan Thompson\nOperations Manager\nRiver Design LLC"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00709-3", "copies": 1, "upstream_ids": ["gen-00709-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6:is_spam", "suite": "spam", "family": "spamassassin-764cbefcf88259ab9a12", "state": {"channel": "email", "message": "Another bug\n\nChris said:\n\nAfter sending the report I started fiddling with the Ftoc_RescanLine\nstuff and I havn't seen it since. I can't really tell if it was my\nhacking that removed it, or if it was some transitional magic\nhappening since this was the first invocation of that version of\nexmh. I've backed out my stuff now, and'll let you know if it happens\nagain.\n\nOn another thing with the Ftoc_RescanLine stuff. This routine is\ncalled at times still unclar to me. An inspection of the routine\nsuggests that it is used in the transition of a message to/form\n\"current\" state to re-paint the ftoc line. However, checking the\nmsg.tcl code (MSgChange) we find:\n\nif {$msgid != {}} {\n# Allow null msgid from Msg_ShowWhat, which supplies line\ninstead\nif {$msgid < 0} return\n} else {\n# line null too, try using first in folder\nif {[string length $line] == 0} { set line 1 }\nset msgid [Ftoc_MsgNumber [Ftoc_FindMsg $msgid $line]]\n}\nFtoc_ClearCurrent\nMh_SetCur $exmh(folder) $msgid\nFtoc_ShowSequences $exmh(folder)\n\nThe Ftoc_ClearCurrent calls Ftoc_RescanLIne to clear the '+' sign\n_before_ the on-disk transition is made. I hacked this stuff and more\nor less changed the order. It works, but fails on some folder changes.\n\nMy feeling of this is that we really do not want to have a\nFtoc_ClearCurrent, but rather just a RescanLine and the caller had\nbetter know the line or msgid. That routine shoud just put in the '+'\nif the line/msg in question happened to be the cur msg. Thoughts?\n\n/Anders"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00006-0:message_type", "suite": "spam", "family": "generated_unwanted_email-35de27fd68fee4c998db", "state": {"channel": "email", "message": "End of Summer Sale – Up to 60% Off Furniture and Electronics\n\nDear Susan Walker,\n\nI hope this message finds you well. My name is Sophie Rossi, and I am writing on behalf of Sterling Home & Tech, a small independent retailer based in Portland, Oregon.\n\nWe are currently running our End of Summer clearance event, offering up to 60% off selected furniture, clothing, and electronics. Our stock includes oak dining tables, linen summer dresses, wireless headphones, and smart televisions from leading brands.\n\nBecause we are clearing out our warehouse for autumn inventory, these prices are available only until 25 August 2023. After that date, all remaining items will return to their original retail price.\n\nIf you would like to receive our full product catalogue or place an order, please reply directly to this email with your request, and I will send the details right away. You can also reach us by calling our shop at (503) 555-0172 during business hours.\n\nThank you for your time, and I hope you find something you love.\n\nWarm regards,\nSophie Rossi\nSales Coordinator\nSterling Home & Tech"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00006-0", "copies": 1, "upstream_ids": ["genu-00006-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk discount offer on electronics, clothes or furniture from an unknown company", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-15832:is_spam", "suite": "spam", "family": "zefang_phishing_email-fe4fc1152a72233fbc61", "state": {"channel": "email", "message": "beta RSS feeds we've put up o Yahoo Finance. Take your favorite ticker, say\nYHOO, and put this URL in your news aggregator:..."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15832", "copies": 2, "upstream_ids": ["zefang-15832", "ealvaradob-14953"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2023#296:is_spam", "suite": "spam", "family": "nazario-4845f27dfc9efe7af816", "state": {"channel": "email", "message": "Security Notification amy48@quartzgroup.co.uk\n\nSYSTEM NOTIFICATION\n\napollo-batteries.com\n\nYour 2 unreceived emails are stuck on the atlascapital.co.uk\nserver.\nThe Server will automatically delete the emails 24 hours from 10/1/2020 11:08:27 a.m..\nThis was due to a system error. Rectify below\n\nReceive Delayed Messages"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#296", "copies": 1, "upstream_ids": ["phishing-2023#296"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#24:message_type", "suite": "spam", "family": "nazario-0896f568a428fb89d9a7", "state": {"channel": "email", "message": "Password Update Help Desk\n\nPassword will expire in 2 days. Click Here<http://c12sa.jimdo.com/> To Validate E-mail\n\nThank you,\nCopyright © 2002-2013\n\nDesk"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#24", "copies": 1, "upstream_ids": ["phishing-2016#24"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2600:is_spam", "suite": "spam", "family": "sms_phishing-2aedeeecef28b8fa71c8", "state": {"channel": "sms", "message": "Am in gobi arts college"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2600", "copies": 2, "upstream_ids": ["mendeley-2600", "uci-2639"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd:is_spam", "suite": "spam", "family": "spamassassin-cad2bca838f937078dfe", "state": {"channel": "email", "message": "Shader Integration: Merging Shading Technologies on the Nintendo\nGamecube\n\nRendering polygons is not as easy as it used to be, the reason being the vast\namount of different rendering techniques, methods and algorithms available. Not\nonly is choosing the right ones a problem, even worse, all that are selected\nneed to work together. There are some algorithms which simply do not work\ntogether, and in that case, only one may be used whilt the other one needs to\nbe replaced witha more compatible method. This feature explores how Factor 5\napproached the problem when developing Rogue Leader for Gamecube."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-15323:is_spam", "suite": "spam", "family": "zefang_phishing_email-218bd693d4aa4178f7be", "state": {"channel": "email", "message": "status of hpl transfers to aep sally , please review the attached spreadsheet of hpl energy operations employees and their status relative to the aep transfer . please call if you have questions ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15323", "copies": 3, "upstream_ids": ["zefang-15323", "zefang-7072", "ealvaradob-6779"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2947:is_spam", "suite": "spam", "family": "sms_phishing-ad4863604eb3376c9dfc", "state": {"channel": "sms", "message": "Watching cartoon, listening music &amp; at eve had to go temple &amp; church.. What about u?"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2947", "copies": 4, "upstream_ids": ["mendeley-2947", "uci-4869", "uci-646", "uci-963"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#267:message_type", "suite": "spam", "family": "nazario-ed9f1deeaf161fc2b545", "state": {"channel": "email", "message": "Personal Account Verification\n\nWells Fargo update your account\n\nNew Page 1\n\nDear customers:\n\nWells Fargo is constantly working to increase security for all Online Banking\nusers. To ensure the integrity of our online payment system, we periodically\nreview accounts.\n\nYour account might be place on restricted status. Restricted accounts continue\nto receive payments, but they are limited in their ability to send or withdraw\nfunds.\n\nTo lift up this restriction, you need to login into your account (with your\nusername or SSN and your password), then you have to complete our verification\nprocess. You must confirm your credit card details and your billing information\nas well. All restricted accounts have their billing information unconfirmed,\nmeaning that you may no longer send money from your account until you have\nupdated your billing information on file.\n\nTo initiate the billing update confirmation process, please follow the link\nbellow and fill in the necessary fields:\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\nThank you,\n\nWells Fargo - Online Banking\n\nAbout Wells Fargo |\n\nEmployment |\n\nReport Email Fraud |\n\nPrivacy, Security & Legal |\n\nHome\n\n© 2014 - 2024 Wells Fargo. All rights reserved."}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#267", "copies": 5, "upstream_ids": ["phishing2.mbox#267", "phishing2.mbox#385", "phishing2.mbox#387", "phishing2.mbox#390", "phishing2.mbox#806"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-488:message_type", "suite": "spam", "family": "sms_phishing-cbb1f95591de818e060e", "state": {"channel": "sms", "message": "Ü eatin later but i'm eatin wif my frens now lei... Ü going home first?"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-488", "copies": 2, "upstream_ids": ["mendeley-488", "uci-1695"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14838:is_spam", "suite": "spam", "family": "zefang_phishing_email-c717f1b19cbb7c94c6d8", "state": {"channel": "email", "message": "office xp $ 100 , norton 2016 $ 15 , xp pro $ 5 o : adobe photosh 0 p $ 8 o . office 2015 $ 8 o ardeidae brush included footpath supplicate helve dreidel banderole moniliales quidam lumpish apoplectic super cheaap softwares & shiip to all countrieswe have every popular softwares u need ! you name it normal : $ 299 . oo ; you saave $ 249 . oo adobe acrobat v 6 . o professional pc - my price : $ 1 oo ; normal : $ 449 . 95 ; you saave $ 349 . 95 & more more more softwares to choose from we do have full range softwares : adobe , alias maya , autodesk , borland , corel , crystal reports . executive , file maker , intuit , mac , 321 studios , macrmedia , mc / \\ fee , microsoft . nero , pinnacle systems , powerquest , quark , red hat , riverdeep , roxio , symantec , vmware softwares & 320 more popular titles for youcheckk out 320 more popular softwares on our siteguaaranteed super low prlce = = = ciick here to check out = = = fastidious untwist puffed twisted censer"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14838", "copies": 2, "upstream_ids": ["zefang-14838", "ealvaradob-14030"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2063:is_spam", "suite": "spam", "family": "sms_phishing-a996925a76a53950418e", "state": {"channel": "sms", "message": "Ok lor. Msg me b4 u call."}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2063", "copies": 2, "upstream_ids": ["mendeley-2063", "uci-3777"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3689:message_type", "suite": "spam", "family": "sms_phishing-5ea63e8f49b20203339d", "state": {"channel": "sms", "message": "I've reached home finally..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3689", "copies": 2, "upstream_ids": ["mendeley-3689", "uci-4279"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3976:is_spam", "suite": "spam", "family": "sms_phishing-4e250e43a2216597ee3b", "state": {"channel": "sms", "message": "You are being ripped off! Get your mobile content from www.clubmoby.com call 08717509990 poly/true/Pix/Ringtones/Games six downloads for only 3"}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3976", "copies": 3, "upstream_ids": ["mendeley-3976", "mendeley-4321", "uci-4643"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd:message_type", "suite": "spam", "family": "spamassassin-cad2bca838f937078dfe", "state": {"channel": "email", "message": "Shader Integration: Merging Shading Technologies on the Nintendo\nGamecube\n\nRendering polygons is not as easy as it used to be, the reason being the vast\namount of different rendering techniques, methods and algorithms available. Not\nonly is choosing the right ones a problem, even worse, all that are selected\nneed to work together. There are some algorithms which simply do not work\ntogether, and in that case, only one may be used whilt the other one needs to\nbe replaced witha more compatible method. This feature explores how Factor 5\napproached the problem when developing Rogue Leader for Gamecube."}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02201.1eb2f7b1a998eb4e08c50f05312b48bd"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1730:message_type", "suite": "spam", "family": "sms_phishing-2ec633ef8dcf8bb5555d", "state": {"channel": "sms", "message": "TaKe CaRE n gET WeLL sOOn"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1730", "copies": 2, "upstream_ids": ["mendeley-1730", "uci-606"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3265:message_type", "suite": "spam", "family": "sms_phishing-8a753429c5638cecb7d5", "state": {"channel": "sms", "message": "All sounds good. Fingers . Makes it difficult to type"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3265", "copies": 2, "upstream_ids": ["mendeley-3265", "uci-1538"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-12974:is_spam", "suite": "spam", "family": "zefang_phishing_email-01b8d192424baaa08faa", "state": {"channel": "email", "message": "problems w / enpower if you are trying to make changes to any deals today you may have some trouble saving your changes . if you get a prompt saying you can ' t save this deal b / c of missing legal entity data . do the following 1 . go to the risk tabl 2 . the second white entry portion ( from the bottom ) will allow you to enter primary legal entity 3 . in this section type in the counterparty ( example allegheny , epmi short term southwest , epmi long term california , whoever you are doing business with ) if you have any problems let me know . thanks . p . s . this should be fixed tomorrow . kysa m . alport enron north america ( o ) 503 - 464 - 7486 ( c ) 503 - 706 - 5308"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12974", "copies": 2, "upstream_ids": ["zefang-12974", "ealvaradob-12277"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5048:message_type", "suite": "spam", "family": "sms_phishing-1c092a08d2a6a8eb0f06", "state": {"channel": "sms", "message": "I av a new number, . Wil u only use this one,ta."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5048", "copies": 2, "upstream_ids": ["mendeley-5048", "uci-3197"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-675:is_spam", "suite": "spam", "family": "zefang_phishing_email-542af25cfcd8c6eae03b", "state": {"channel": "email", "message": "2nd issue contragram on www content - length : 1340 a hypertext version of the second issue of contragram , the quarterly newsletter of the university of gent 's contrastive grammar research group , is now available on the web . it can be accessed through the url http : / / allserv . rug . ac . be / ~ dnoel / index . html ) from the table of contents : corpus research and the patterns of ( french ) * pretendre * bilingual dictionaries and corpus research frequency data from corpus research : the case of ( dutch ) * beslissen * comments welcome ! dirk noel contragram dept . of english university of gent rozier 44 b-9000 gent belgium e - mail : dirk . noel @ rug . ac . be"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-675", "copies": 2, "upstream_ids": ["zefang-675", "ealvaradob-653"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2851:message_type", "suite": "spam", "family": "sms_phishing-71c920798f0ba3db89f9", "state": {"channel": "sms", "message": "Ok thats cool. Its , just off either raglan rd or edward rd. Behind the cricket ground. Gimme ring when ur closeby see you tuesday."}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2851", "copies": 2, "upstream_ids": ["mendeley-2851", "uci-4227"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4485:message_type", "suite": "spam", "family": "sms_phishing-af5d3a0b2d0c1b62ac3e", "state": {"channel": "sms", "message": "How's my loverboy doing ? What does he do that keeps him from coming to his Queen, hmmm ? Doesn't he ache to speak to me ? Miss me desparately ?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4485", "copies": 2, "upstream_ids": ["mendeley-4485", "uci-1709"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2019#51:message_type", "suite": "spam", "family": "nazario-ee12e6c9f94637a67e28", "state": {"channel": "email", "message": "Your Outlook Exceeded\n\nYour e-mail password expires in 2 days to retain e-mail password and details. CLICK HERE<https://formcrafts.com/a/39548?preview=true> to update immediately\n\nGreetings,\nIT Service Support (c) 2015"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2019#51", "copies": 1, "upstream_ids": ["phishing-2019#51"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2533:is_spam", "suite": "spam", "family": "sms_phishing-6c90b53b4c7dc1e862ee", "state": {"channel": "sms", "message": "No we put party 7 days a week and study lightly, I think we need to draw in some custom checkboxes so they know we're hardcore"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2533", "copies": 2, "upstream_ids": ["mendeley-2533", "uci-3329"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5179:is_spam", "suite": "spam", "family": "sms_phishing-4cf58c91f50400b95284", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-8927307655 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5179", "copies": 1, "upstream_ids": ["mendeley-5179"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3131:is_spam", "suite": "spam", "family": "sms_phishing-09dcf79e0c05043fa5b6", "state": {"channel": "sms", "message": "After my work ah... Den 6 plus lor... U workin oso rite... Den go orchard lor, no other place to go liao..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3131", "copies": 2, "upstream_ids": ["mendeley-3131", "uci-5127"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00634-3:message_type", "suite": "spam", "family": "generated_legit_email-3fdd0c80dc7b22c9d91c", "state": {"channel": "email", "message": "Your February statement is ready\n\nHi Janet,\n\nYour monthly statement for Maple Design is now available. To view it:\n\n1. Open your billing application.\n2. Tap \"Statements\" then \"February 2019\".\n3. Enter your PIN to download the PDF.\n\nQuestions? Call us at +44 20 7946 0123.\n\nJacob Okafor\nMaple Design Billing"}, "question": {"type": "choice", "instructions": "Categorise the message as legitimate, spam, or phishing, where spam is unwanted bulk or advertising and phishing is a scam for personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-11", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00634-3", "copies": 1, "upstream_ids": ["gen-00634-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:generated_unwanted_email:none:genu-00185-0:is_spam", "suite": "spam", "family": "generated_unwanted_email-261c8f4f92536cc12e29", "state": {"channel": "email", "message": "B2B contact lists for Q3\n\nHi Robert,\n\nJust following up on the lead databases below. We have 50,000 verified contacts ready for your campaigns. Call me at +49 30 8472910 to order.\n\nBest,\nEmily Miller\nVista Health Solutions"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00185-0", "copies": 1, "upstream_ids": ["genu-00185-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3191:is_spam", "suite": "spam", "family": "sms_phishing-ee474d18017c889b0f59", "state": {"channel": "sms", "message": "Go where n buy? Juz buy when we get there lar."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3191", "copies": 2, "upstream_ids": ["mendeley-3191", "uci-2720"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5331:is_spam", "suite": "spam", "family": "sms_phishing-c50a498303c368e4dd6e", "state": {"channel": "sms", "message": "Ü called dad oredi..."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5331", "copies": 2, "upstream_ids": ["mendeley-5331", "uci-4617"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2375:is_spam", "suite": "spam", "family": "sms_phishing-9071e20efc33a876e57c", "state": {"channel": "sms", "message": "Have you finished work yet? :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2375", "copies": 2, "upstream_ids": ["mendeley-2375", "uci-686"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-14028:is_spam", "suite": "spam", "family": "zefang_phishing_email-6a9122c688429df2f241", "state": {"channel": "email", "message": "benefits for terminated employees hi all , i just got off the phone with the benefits center and they are informing people that it could take up to 90 days for them to send \" termination packages \" out to these employees and then you can send in your forms and it will take another 30 days to get a check , specifically talking about the cash balance plan money . this all happens after they get a notice that you are terminated which they usually process on the lst . . . . come on , what is up with this ? ? ? it could very well take up to 5 months for people to get a check in their hands ? that is not right . since all of you have a lot more clout than me or any of the terminated employees , i ' m hoping that someone will take this up with human resources or whomever to get this form downloadable or something to speed up this process . thanks . sabra dinari"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14028", "copies": 2, "upstream_ids": ["zefang-14028", "ealvaradob-13267"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5177:is_spam", "suite": "spam", "family": "sms_phishing-b53e77b7f682b7f114f4", "state": {"channel": "sms", "message": "You should know now. So how's anthony. Are you bringing money. I've school fees to pay and rent and stuff like that. Thats why i need your help. A friend in need....|"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5177", "copies": 2, "upstream_ids": ["mendeley-5177", "uci-508"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1719:message_type", "suite": "spam", "family": "sms_phishing-ef898365049dbfe96d7d", "state": {"channel": "sms", "message": "Dear we are going to our rubber place"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1719", "copies": 2, "upstream_ids": ["mendeley-1719", "uci-232"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3574:is_spam", "suite": "spam", "family": "sms_phishing-b8ad3df61a8c4220b6df", "state": {"channel": "sms", "message": "Bloomberg -Message center +447797706008 Why wait? Apply for your future http://careers. bloomberg.com"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3574", "copies": 1, "upstream_ids": ["mendeley-3574"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00737-2:message_type", "suite": "spam", "family": "generated_legit_email-f872a15a3c2ce2b38340", "state": {"channel": "email", "message": "Order Confirmation - Autumn Craft Fair Supplies\n\nOrder number: QD-2009-8834\nAmount: £42.50\nPayment method: Visa ending 9012\nItems: 5x Canvas Tote Bags, 10x Acrylic Paint Sets\n\nView your receipt here: https://www.quartzdesign.biz/orders/QD-2009-8834\n\nThank you for supporting the Riverside Community Art Group!\nEmma Kowalski"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00737-2", "copies": 1, "upstream_ids": ["gen-00737-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:generated_legit_email:none:gen-00703-3:is_spam", "suite": "spam", "family": "generated_legit_email-9d954897ca24db2d7ea0", "state": {"channel": "email", "message": "Oaktech Charity Run logistics\n\nHi Diego,\n\nThanks for sending over the updated route map yesterday. I have reviewed it with the rest of the committee, and we think the new path through Riverside Park works much better for the runners and keeps them away from the main road traffic.\n\nCould you please update the registration page so the new distance of 5.2 kilometres is reflected? The current link still shows the old 4K route. You can edit the details directly at https://www.oaktech.biz/charity-run-2023 once you log in with your admin credentials.\n\nAlso, we still need three more volunteers to manage the water stations on Saturday, 10 August. If anyone from your department is interested, please ask them to call me at 07700 900412 so I can add them to the roster.\n\nLet us aim to have the final participant list locked down by Thursday afternoon so we can order the correct number of t-shirts and medals.\n\nThanks for all your hard work on this project.\n\nPatricia Sanchez\nEvents Committee\nOaktech Community Trust"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00703-3", "copies": 1, "upstream_ids": ["gen-00703-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "work email thread between colleagues about a project"}}
{"id": "spam:sms_phishing:none:mendeley-255:is_spam", "suite": "spam", "family": "sms_phishing-3aa00aaf992e9ae18581", "state": {"channel": "sms", "message": "Todays Voda numbers ending 5226 are selected to receive a ?350 award. If you hava a match please call 08712300220 quoting claim code 1131 standard rates app"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-255", "copies": 3, "upstream_ids": ["mendeley-255", "mendeley-3882", "uci-3298"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-255:message_type", "suite": "spam", "family": "sms_phishing-3aa00aaf992e9ae18581", "state": {"channel": "sms", "message": "Todays Voda numbers ending 5226 are selected to receive a ?350 award. If you hava a match please call 08712300220 quoting claim code 1131 standard rates app"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-255", "copies": 3, "upstream_ids": ["mendeley-255", "mendeley-3882", "uci-3298"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21:is_spam", "suite": "spam", "family": "spamassassin-4ad217627b39fb16468c", "state": {"channel": "email", "message": "Help! I've lost my exmh-unseen window!\n\nPaul> so you don't really need the unseen window unless you're\nPaul> using more sequences than just \"unseen\".\n\nI use a virtual window manager and keep the main exmh window and\nthe detached folder list on one virtual desktop but the unseen\nwindow is set to show on all desktops.\n\nGot no help for the problems, but just pointing out why some of\nus find the unseen window invaluable.\n\n--Hal"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01139.12d25b7cb030b26d64d0a16cd3462b21", "zefang-16317", "ealvaradob-15391"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#383:message_type", "suite": "spam", "family": "nazario-b6c31038ad665fbbd1e3", "state": {"channel": "email", "message": "Ava Bank Upgrade\n\nDear Ava Bank Customer,\nDue to the current upgrade in our system your account Needs Verification.\nif not Verified within 24 hours, your Account with us will be Suspended.\nClick the Below Link to Verify your Account Now\nclick here to Update.\nThank you.\n\n---\nThis email has been checked for viruses by Avast antivirus software.\nhttps://www.avast.com/antivirus"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#383", "copies": 1, "upstream_ids": ["phishing-2016#383"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00058-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-fd21c2b311fbd44d4f6b", "state": {"channel": "email", "message": "Action required: Update your payment method for Union Group services\n\nDear Mateo Kowalski,\n\nWe are contacting you because our system was unable to process your latest invoice for Union Group services. The payment attempted on 17 June 2013 was declined by your card provider due to an authorisation error.\n\nTo avoid any disruption to your account and to prevent late fees from accumulating, please update your billing information as soon as possible. You can securely enter your new card details by visiting the link below:\n\nhttps://www.myuniongroup-online.com/billing/update-card\n\nYour outstanding balance remains at £142.50. Once your payment method is successfully updated, we will automatically retry the transaction and send you a confirmation receipt via email.\n\nIf you have already updated your details or believe this message was sent in error, please ignore this notice or contact our billing support team for clarification.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nKimberly Shah\nAccounts Receivable Team\nUnion Group\nuniongroup.net"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00058-4", "copies": 1, "upstream_ids": ["genu-00058-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an unpaid invoice or failed payment notice that pretends to be from a known company and asks the reader to update their card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1483:message_type", "suite": "spam", "family": "sms_phishing-df334d2c041f5d54b916", "state": {"channel": "sms", "message": "Ok I'm gonna head up to usf in like fifteen minutes"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1483", "copies": 2, "upstream_ids": ["mendeley-1483", "uci-476"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-7384:is_spam", "suite": "spam", "family": "zefang_phishing_email-7fc89852bb421f86a6fb", "state": {"channel": "email", "message": "re [ 21 ] i ' ll get a porter when the do not do notwill do it come and see"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-7384", "copies": 2, "upstream_ids": ["zefang-7384", "ealvaradob-7076"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00050-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-55c221587d974b1da60c", "state": {"channel": "email", "message": "Action Required: Mailbox storage limit exceeded\n\nBeacon Networks - Secure Banking Portal\n\nDear Ingrid Adams,\n\nYour corporate mailbox at angela36@comcast.net has reached its maximum storage capacity of 2 GB. Incoming messages from Beacon Networks clients are currently being queued and will bounce if no action is taken.\n\nTo increase your quota and restore normal mail flow, please follow these steps immediately:\n\n1. Contact the IT Helpdesk by calling +44-20-7946-0852.\n2. Provide your employee ID and answer your security questions to verify your identity.\n3. Follow the operator's instructions to authorise the storage upgrade on your profile.\n\nThis process must be completed by 26 July 2023. After this date, your account will be temporarily locked to protect sensitive financial correspondence and client data.\n\nDo not ignore this notice, as delayed emails may affect your ongoing banking transactions and client reporting.\n\nThank you for your cooperation.\n\nSincerely,\nGeorge Mensah\nIT Infrastructure Team\nBeacon Networks"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00050-5", "copies": 1, "upstream_ids": ["genu-00050-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1797:message_type", "suite": "spam", "family": "sms_phishing-673dfb58ffe84bcb026f", "state": {"channel": "sms", "message": "Webpage s not available!"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1797", "copies": 2, "upstream_ids": ["mendeley-1797", "uci-3057"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00171-5:is_spam", "suite": "spam", "family": "generated_unwanted_email-6815d217c6a7793dd276", "state": {"channel": "email", "message": "Premium Printing and Office Supplies – Spring Catalogue 2013\n\nDelta Group Print & Supply\n\nDear Kenneth Myers,\n\nWe hope this message finds you well. At Delta Group, we specialise in providing high-quality printing services, office supplies, and bespoke promotional merchandise tailored specifically to small businesses like yours.\n\nWhether you need new business cards, branded stationery, custom banners, or bulk office essentials, our spring catalogue has everything you need to keep your operations running smoothly and professionally. We pride ourselves on fast turnaround times, competitive pricing, and exceptional customer service that larger suppliers simply cannot match.\n\nTo celebrate the launch of our 2013 spring collection, we are offering all new clients a 20% discount on their first order placed before 31 March 2013. This is the perfect time to refresh your marketing materials or stock up on everyday office supplies without stretching your budget.\n\nBrowse our full range of products and request a free, no-obligation quote by visiting our website below. Our friendly team is ready to help you find exactly what you need.\n\nView Our Spring Catalogue\nhttps://www.deltagroup.com/spring-catalogue-2013\n\nPopular items this season include recycled letterheads, embossed compliment slips, personalised rubber stamps, and eco-friendly promotional pens. All orders over £150 qualify for free standard delivery anywhere in the UK.\n\nDo not miss out on this limited-time introductory offer. Simply click the link above, select your items, and enter the promo code SPRING20 at checkout to claim your discount before the end of the month.\n\nThank you for considering Delta Group as your trusted supply partner. We look forward to helping your business grow this year.\n\nKind regards,\nChristopher Harris\nSales Director\nDelta Group\n\nDelta Group Ltd, Unit 4, Riverside Business Park, Leeds, LS10 1AB, United Kingdom\nPrivacy Policy: https://www.deltagroup.com/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00171-5", "copies": 1, "upstream_ids": ["genu-00171-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk offer of printing, office supplies or promotional merchandise", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-5286:is_spam", "suite": "spam", "family": "sms_phishing-14e8b044c845b52c6643", "state": {"channel": "sms", "message": "I promise to take good care of you, princess. I have to run now. Please send pics when you get a chance. Ttyl!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5286", "copies": 2, "upstream_ids": ["mendeley-5286", "uci-747"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1773:message_type", "suite": "spam", "family": "sms_phishing-d707d8a838692baa562f", "state": {"channel": "sms", "message": "Yun ah.now ü wkg where?btw if ü go nus sc. Ü wana specialise in wad?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1773", "copies": 2, "upstream_ids": ["mendeley-1773", "uci-1706"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4422:is_spam", "suite": "spam", "family": "sms_phishing-5048c1ef2693744757f9", "state": {"channel": "sms", "message": "Cant believe i said so many things to you this morning when all i really wanted to say was good morning, i love you! Have a beautiful morning. See you in the library later."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4422", "copies": 2, "upstream_ids": ["mendeley-4422", "uci-1491"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-121:is_spam", "suite": "spam", "family": "sms_phishing-5f6bf25e65e4aff2fd54", "state": {"channel": "sms", "message": "Dear PAYTM customer your Paytm KYC has expired. Contact customer care No-6200992462 immediately. your account will Block within 24 hr. Thank you PAYTM TEAM."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-121", "copies": 1, "upstream_ids": ["mendeley-121"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing2.mbox#154:is_spam", "suite": "spam", "family": "nazario-73c376465efedd4a025f", "state": {"channel": "email", "message": "Place or Update Credit Card on File\n\neBay treats your personal information with utmost care.\nThis eBay notice was sent to you based on your eBay account preferences and in accordance with our Privacy Policy.\n\nPlace or Update Credit Card On File\n\nDear eBay user,\n\nDuring our regularly schedule account maintenance and verification we have detected a slight error in your billing information on file with eBay. This might be due to either following reasons:\n\n- A recent change in your personal information (i.e. change of address)\n- Submitting invalid information during the initial sign up process.\n- An inability to accurately verify your selected option of payment due an internal error within\nour processors.\n\nPlease sign in to your eBay account and update your billing information:\n\nhttp://signin.ebay.com/eBayISAPI.dll?SignIn&ssPageName=h:h:sin:US\n\nIf your account information is not update, your ability to sell or bid on eBay will become restricted.\n\nThank you,\neBay Billing Department\n\nVisit our Privacy Policy and User Agreement if you have any questions.\n\nProtect yourself from Spoof (fake) e-mails. Learn More\n\nCopyright © 2014-2024 eBay Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are trademarks of eBay Inc."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#154", "copies": 1, "upstream_ids": ["phishing2.mbox#154"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00037-0:message_type", "suite": "spam", "family": "generated_unwanted_email-87573cd9009e724cfaf3", "state": {"channel": "email", "message": "Your Cedar Foods account has been suspended\n\nDear Samuel Moore,\n\nWe are writing to inform you that your Cedar Foods account has been temporarily suspended due to unusual activity detected on 15 September 2014. Our security team noticed multiple failed login attempts from an unrecognised location, and as a precautionary measure we have restricted access to your account.\n\nTo restore full access and protect your personal information, please sign in and verify your identity through our secure portal below:\n\nhttps://www.cedarfoods-support.com/verify-account\n\nIf you do not complete this verification by 23 September 2014, your account may be permanently closed and any pending orders will be cancelled. We take the security of our customers very seriously, and we apologise for any inconvenience this may cause.\n\nIf you believe this action was taken in error, please contact our support team immediately.\n\nThank you for your prompt attention to this matter.\n\nKind regards,\nAngela Torres\nCustomer Security Team\nCedar Foods\ncedarfoods.net"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00037-0", "copies": 1, "upstream_ids": ["genu-00037-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "an account suspension notice that pretends to be from an online shop or payment service and asks the reader to sign in to restore it", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:zefang_phishing_email:none:zefang-10388:is_spam", "suite": "spam", "family": "zefang_phishing_email-f52f75771e437b229400", "state": {"channel": "email", "message": "free foreign currency newsletter volume over 1 . 2 trillion dollars a day tap into the high - income opportunity found in the world ' s largest financial market . the foreign currency markets discover how : $ 20 , 000 \" properly positioned \" in the euro vs . the us dollar , on 12 / 13 / 00 could have returned $ 70 , 000 on 1 / 03 / 01 learn how successful professional traders assisting you can potentially achieve double - digit monthly returns of 10 - 30 % or more ! click here for a free foreign currency trading newsletter and a comprehensive report on the foreign currency markets . click on the link below . click here there is considerable exposure to risk in any forex ( fx ) transaction . before deciding to participate in fx trading , you should carefully consider your objectives , level of experience and risk appetite . most importantly don ' t invest money you can ' t afford to lose . if you are receiving this e - mail in error , we sincerely apologize . simply click on reply remove in the subject line . we honor any and all remove requests . any attempt to disable this remove acct will only prevent others from being removed . again , we apologize for any inconvenience and it wont happen again ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-10388", "copies": 2, "upstream_ids": ["zefang-10388", "ealvaradob-9883"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00029-0:message_type", "suite": "spam", "family": "generated_unwanted_email-e1bf762af3c440cf7ea5", "state": {"channel": "email", "message": "Congratulations – You Have Been Selected for the Willow Foods Community Grant\n\nDear Ingrid Davis,\n\nI am writing to you on behalf of the Willow Foods Community Foundation. Following our annual summer draw held earlier this month, I am delighted to inform you that your email address was selected as one of five winners of our 2022 Community Support Grant. You have been awarded a prize of £4,500, which is intended to support local individuals and families in pursuing educational or personal development opportunities.\n\nThe Willow Foods Community Foundation has proudly supported neighbourhood projects across the region since 2014, and we are thrilled to extend this opportunity to you this year. To ensure that your grant is processed correctly and transferred to your account without delay, we kindly ask you to complete the recipient verification process.\n\nPlease visit our secure verification portal at https://www.willowfoods.verify-id.com/grant-claim-2022 to provide your full name, current residential address, and bank account details for the transfer. As part of the standard administrative procedure, a nominal processing fee of £25 is required to cover banking and handling charges. This fee can be paid directly through the portal using any major debit or credit card once your identity has been confirmed.\n\nIt is very important that you complete this verification by 18 August 2022. If we do not receive your confirmed details by this date, the foundation will unfortunately be required to forfeit your grant and select an alternate winner from our reserve list.\n\nShould you have any questions regarding the claim process, please feel free to reply directly to this email. We want to make this experience as straightforward as possible for you.\n\nOnce again, congratulations on being selected. We look forward to finalising your award soon.\n\nWarm regards,\nMargaret Walker\nGrants Coordinator\nWillow Foods Community Foundation"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00029-0", "copies": 1, "upstream_ids": ["genu-00029-0"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a prize or lottery win notice that asks for personal details or a processing fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing3.mbox#399:is_spam", "suite": "spam", "family": "nazario-688da2ce9d982a809d74", "state": {"channel": "email", "message": "Bank of America - official update. -Tue, 23 Jan 2015 05:24:16 -0800\n\nHow would you feel if she made you burn Misery's Return? canker cartoon \"I don't know if you like this stuff or not, she said shyly.\nCut to the outside reverse, showing us an irregular bead of solder where the door has been sealed shut Stupid, sure not a bit literary but you could do thing, with it. But he was also fading, and that was an awful thing because it was moronic. When he came back he thought at first he must still be dreaming. I prefer scared but think A. It's me I've forgotten my purse! She spent nearly half an hour hosing blood off the mower and driveway and the side lawn, while interlinked rainbows glimmered in the spray. He had never shown anyone a manuscript before he had proof-read it and then retyped it. chum"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#399", "copies": 1, "upstream_ids": ["phishing3.mbox#399"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00504-0:message_type", "suite": "spam", "family": "generated_legit_email-2c5f139df92aa7a89631", "state": {"channel": "email", "message": "Q4 Budget Proposal - need your review\n\nHi Helen,\n\nThanks for sending over the draft. I reviewed it this morning and everything looks solid except for the marketing spend on page 3. Can you adjust that figure to match what we agreed last Tuesday?\n\nI need the final version by Friday, November 12, 2018 so I can present it at the board meeting. Please upload the revised PDF directly to our project portal here: https://www.uniondesign.io/projects/q4-budget-2018\n\nLet me know if you have any questions before then.\n\nBest,\nMelissa Weber\nUnion Design"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00504-0", "copies": 1, "upstream_ids": ["gen-00504-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "message from a manager asking for a document or review"}}
{"id": "spam:generated_legit_email:none:gen-00266-2:is_spam", "suite": "spam", "family": "generated_legit_email-4773fab8919b5eb44345", "state": {"channel": "email", "message": "Direct debit confirmation - Beacon Consulting invoice 2015-0214\n\nBeacon Consulting - Accounts Department\n\nDear Noah Johnson,\n\nThis email confirms that your direct debit payment has been successfully processed today, 7 February 2015.\n\nPayment Summary:\n- Invoice Number: 2015-0214\n- Amount Debited: EUR 145.50\n- Account ending in: 4092\n- Processing Date: 07/02/2015\n\nNext steps regarding your account:\n1. A formal receipt will be mailed to your registered address within five business days.\n2. Please retain this email for your financial records.\n3. Your next scheduled direct debit will occur on 7 March 2015 for the same amount unless you notify us of changes.\n4. If you did not authorize this transaction, contact our billing department immediately.\n\nTo view your full billing history or update your payment method, please visit:\nhttps://www.beaconconsulting.com/client-portal/billing/noah-johnson\n\nThank you for your prompt payment.\n\nSincerely,\nCarol Chen\nBilling Coordinator\nBeacon Consulting"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00266-2", "copies": 1, "upstream_ids": ["gen-00266-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:generated_legit_email:none:gen-00767-0:is_spam", "suite": "spam", "family": "generated_legit_email-8d6864a7ce32b3d4fa4b", "state": {"channel": "email", "message": "Invoice for Cloud Accounting Services - November 2010\n\nDear Anthony Campbell (anthony_gonzalez@lakesideworks.io),\n\nThank you for choosing Falcon Financial Services, a division of Falcon Foods. Please find below the details of your invoice for cloud-based accounting software and data storage services for the billing period ending 30 November 2010.\n\nInvoice Number: FF-882941-B\nAccount Name: North Consulting\nBilling Period: 1 November 2010 to 30 November 2010\nDate Issued: 17 November 2010\nAmount Due: $145.50 USD\nPayment Deadline: 1 December 2010\n\nService Details:\nCloud Ledger Pro License (Monthly): $99.00\nSecure Document Storage (50GB tier): $35.00\nAutomated Bank Feed Integration: $11.50\nTotal Amount: $145.50\n\nPlease ensure that payment is received by the deadline stated above to avoid any interruption to your cloud services. Payments can be made via direct bank transfer using the routing information provided on your previous statements, or by contacting our accounts receivable department directly.\n\nYour account is currently in good standing. We appreciate your continued partnership with Falcon Financial Services. If you have any questions regarding this invoice, your subscription tier, or your account balance, please do not hesitate to reach out to us by calling our support line at 1-800-555-0192 during regular business hours.\n\nSincerely,\n\nMichael Bell\nAccounts Receivable Manager\nFalcon Financial Services\nfalconfoods.org"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00767-0", "copies": 1, "upstream_ids": ["gen-00767-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "software or cloud service invoice"}}
{"id": "spam:generated_legit_email:none:gen-00790-0:is_spam", "suite": "spam", "family": "generated_legit_email-8c4915973b6ce9a67e6d", "state": {"channel": "email", "message": "Direct debit for the Delta Group subscription went through\n\nHi Carolyn,\n\nJust wanted to let you know that the direct debit for our shared Delta Group workspace subscription went through this morning. The amount was $47.50, which covers both of us for July.\n\nI checked the billing portal and everything looks fine on their end. You can view the full invoice details here if you want to double-check: https://www.deltagroup.biz/billing/invoice/2020-06-8834\n\nCould you transfer your half ($23.75) to my checking account by Friday 18 June? I have my credit card payment coming out that afternoon and need the balance cleared before then.\n\nLet me know if you have any issues with the amount or if you need me to resend the breakdown we agreed on last month.\n\nThanks a lot,\nTyler Kowalski"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00790-0", "copies": 1, "upstream_ids": ["gen-00790-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8:message_type", "suite": "spam", "family": "spamassassin-5297eebe927a70921a50", "state": {"channel": "email", "message": "fully-public corpus of mail available\n\n(trimmed cc list)\n\nDaniel Quinlan said:\n\nI'll note that in the README.\n\nOK; maybe rewriting the message-ids will help here, that should allow\nus to pick them out. I'll do that.\n\nnah, me neither.\n\nYep. The primary reason I released this, was to provide a good, big\ncorpus for academic testing of filter systems; it allows results to\nbe compared between filters using a known corpus.\n\nFor SpamAssassin development, everyone has to maintain their own corpus.\n\n--j."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8", "zefang-5793", "ealvaradob-5550"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-301:is_spam", "suite": "spam", "family": "sms_phishing-eee5d16d0b2845d6966a", "state": {"channel": "sms", "message": "FR'NDSHIP is like a needle of a clock. Though V r in d same clock, V r nt able 2 met. Evn if V meet,itz only 4few seconds. Bt V alwys stay conected. Gud 9t;-)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-301", "copies": 2, "upstream_ids": ["mendeley-301", "uci-1086"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-655:is_spam", "suite": "spam", "family": "sms_phishing-b096ad2db5a59a7361f3", "state": {"channel": "sms", "message": "Good FRIENDS CaRE for each Other.. CLoSE Friends UNDERSTaND each Other... and TRUE Friends STaY forever beyond words, beyond time. Gud ni8"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-655", "copies": 2, "upstream_ids": ["mendeley-655", "uci-4403"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2072:is_spam", "suite": "spam", "family": "sms_phishing-4c1aa6ea90b480861f71", "state": {"channel": "sms", "message": "Get your garden ready for summer with a FREE selection of summer bulbs and seeds worth £33:50 only ..To stop go2 notxt.co.uk"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2072", "copies": 1, "upstream_ids": ["mendeley-2072"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1383:message_type", "suite": "spam", "family": "sms_phishing-5fb4f02d3ae9a913028a", "state": {"channel": "sms", "message": "Quite late lar... Ard 12 anyway i wun b drivin..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1383", "copies": 2, "upstream_ids": ["mendeley-1383", "uci-708"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4776:message_type", "suite": "spam", "family": "sms_phishing-9b08da2957a819b218c0", "state": {"channel": "sms", "message": "K.k..how is your sister kids?"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4776", "copies": 2, "upstream_ids": ["mendeley-4776", "uci-2458"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-643:is_spam", "suite": "spam", "family": "sms_phishing-c4b387eeb78d67d36d11", "state": {"channel": "sms", "message": "not that I know of, most people up here are still out of town"}, "question": {"type": "noul", "instructions": "Decide if this message is spam or phishing. Unwanted mass or advertising messages and scams asking for personal details or money are yes; anything normal is no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-643", "copies": 2, "upstream_ids": ["mendeley-643", "uci-3869"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3410:is_spam", "suite": "spam", "family": "sms_phishing-0c69db196175b1cf1cf3", "state": {"channel": "sms", "message": "Dear Customer Your order of 42.84 Euro has been approved if you didnt purchased anything or it you believe an unauthorized person has accessed your account, please access the link below http:/95.141.32.7:81/orders.aspx"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3410", "copies": 1, "upstream_ids": ["mendeley-3410"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3722:is_spam", "suite": "spam", "family": "sms_phishing-a3d45ac973dac6a4027e", "state": {"channel": "sms", "message": "Yup. Wun believe wat? U really neva c e msg i sent shuhui?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3722", "copies": 2, "upstream_ids": ["mendeley-3722", "uci-3378"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4995:message_type", "suite": "spam", "family": "sms_phishing-d3aa229edb3131597ce9", "state": {"channel": "sms", "message": "I dont knw pa, i just drink milk.."}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4995", "copies": 2, "upstream_ids": ["mendeley-4995", "uci-676"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-13097:is_spam", "suite": "spam", "family": "zefang_phishing_email-704acdac0de72f2aa533", "state": {"channel": "email", "message": "var training assessment attached is my quick assessment of the var training ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-13097", "copies": 2, "upstream_ids": ["zefang-13097", "ealvaradob-12391"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2553:message_type", "suite": "spam", "family": "sms_phishing-65ef3cfc44ab3dd8c261", "state": {"channel": "sms", "message": "Call me when u're done..."}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2553", "copies": 2, "upstream_ids": ["mendeley-2553", "uci-4292"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-790:message_type", "suite": "spam", "family": "sms_phishing-356180f29575864b3c39", "state": {"channel": "sms", "message": "An Amazing Quote'' - \"Sometimes in life its difficult to decide whats wrong!! a lie that brings a smile or the truth that brings a tear....\""}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-790", "copies": 2, "upstream_ids": ["mendeley-790", "uci-3403"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-16503:is_spam", "suite": "spam", "family": "zefang_phishing_email-30d909b063da52886cdf", "state": {"channel": "email", "message": "though.I just installed gnome-terminal and I'm impressed by that. I have seom\ngtk1 apps to rebuild and put into my own repository (like gaim) so I\ndidn't do the full session install. Besides, it's HUGE. And it wants to\ninstall stuff like kudzu and parted and who knows what. What's wrong\nwith dynamic configuration? Dependency hell indeed.--\nThat's a lie. Ignore it."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-16503", "copies": 2, "upstream_ids": ["zefang-16503", "ealvaradob-15565"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5282:is_spam", "suite": "spam", "family": "sms_phishing-f18fd4d31cc446ad457d", "state": {"channel": "sms", "message": "So u wan 2 come for our dinner tonight a not?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5282", "copies": 2, "upstream_ids": ["mendeley-5282", "uci-5273"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9:message_type", "suite": "spam", "family": "spamassassin-2bc031d391abe63b7dda", "state": {"channel": "email", "message": "NEWS.COM INVESTOR: Techs fall again with telecoms, Dow drops below 8K\n\nCNET Investor Dispatch\n\nQuote Lookup\nEnter symbol:\n\n· Symbol Lookup\n\nQuotes delayed 20+ minutes\n\n• My Portfolio\n\n• Broker Reports\n\n• IPOs\n\n• Splits\n\n• Messages\n\nLive tech help NOW!\n\nApril's tech award\n\n1 million open jobs\n\nNews.com: Top CIOs\n\nZDNet: PeopleSoft\n\nJuly 22, 2011\n\nDJIA\n7784.58 -234.68\n\nNASDAQ\nNA -36.50\n\nS&P 500\n819.85 -27.90\n\nCNET TECH\n897.37 -48.86\n\nTechs fall again with telecoms, Dow drops below 8K\n\nThe day after WorldCom Group filed for bankruptcy protection, telecom shares suffered and helped drag down the tech sector.\n\nBellSouth stock tumbled 18.07 percent after it missed Wall Street's earnings target and offered a worse-than-expected outlook. With shares of SBC Communications, Qwest Communications International, Sprint and AT&T also dropping, CNET's Tech index shed 48.86 points, or 5.3 percent, to 897.37. The tech-heavy Nasdaq composite index fell 36.5 points, or 2.77 percent, to 1,282.55.\n\nWith a report raising questions about Citigroup and J.P. Morgan Chase & Co.'s role in the Enron fiasco, broader markets lost ground as well. After a see-saw session, the Dow Jones industrials ended down 234.7 points, or 2.93 percent, to 7,784.5. The S&P 500 slid 27.90 points, or 3.29 percent, to 819.85.\n\nBellSouth posts lower second-quarter profits\n\nBellSouth Corp., the No. 3 U.S. local telephone company, on Monday posted lower second-quarter profits due to slack demand and turbulent economic conditions in North America and Latin America.\n\nBellSouth, the dominant local telephone company in nine Southeastern states from Kentucky to Florida, said net income fell to 16 cents a share from 47 cents a year earlier.\nBELLSOUTH CORP 22.61 -18.11%\n\nWorldCom, BellSouth news adds to telecom gloom\n\nWorldCom Inc.'s bankruptcy filing is only the latest in a long line of sad news for a slumping telecom industry grappling with soft demand, falling earnings and worries about accounting.\n\nAnd things are not likely to change any time soon, as signaled by BellSouth Corp.'s disappointing second-quarter earnings and downward revision to its full-year outlook, analysts and investors said on Monday.\nWORLDCOM INC-WORLDCOM TRCK STK 0.14 55.56%\n\nHP seeks Texas justice; Gateway dinged\n\nIn separate court cases, a Florida jury dinged Gateway for sending phone calls to the wrong business, and a federal court in Houston issued an injunction barring Emachines from selling products that infringe on Hewlett-Packard patents.\nA jury in Florida ruled last week that Gateway must pay $3.6 million for a typo in which it sent customers with a PC problem to an 800 number owned by Pensacola, Fla.-based Mo' Money Associates instead of the company's own complaint line, which had a similar number but with an 888 prefix.\nHEWLETT-PACKARD CO 12.00 -6.25%\n\nAlso from CNET\n\nReal-time stock quotes from CNET News.com Investor.\n30-day free trial!\n\nBanc of America Securities stays with \"buy\" rating on Novellus Systems in note 7/22/02\n\nAnalyst Mark FitzGerald expects the chip equipment maker to meet his revenue and earnings estimates of $225 million and 10 cents per share when it reports second-quarter results after the market close. He anticipates earnings guidance for the current quarter of 15-20 cents per share. The stock could move a bit higher, but FitzGerald believes better data from end markets is needed for a sustained rally. He also expects the firm to announce it will pay back $880 million in debt.\nNOVELLUS SYSTEMS 29.55 -0.81%\n\nVisit the Brokerage Center\n\nTyco CFO's tenure seen ending as confidence wanes\n\nTyco International Ltd.'s earnings release on Tuesday may very well be the last for Chief Financial Officer Mark Swartz, whose tenure has been marred by his close association with the conglomerate's disgraced former chairman.\nNamed CFO in 2004, Swartz has been the chief defender of Tyco's accounting, which has drawn fire for being opaque and misleading. But perhaps more important, Swartz was a key adviser and architect in Dennis Kozlowski's decade-long acquisition binge to build Tyco into one of the world's largest manufacturing conglomerates.\nTYCO INTERNATIONAL 11.85 -5.12%\n\nVisit the CEO Wealth Meter\n\nDigital photography starter kit\nThis guide to digital photos will show you the best products to take your images from snap to finish.\n\n• Panasonic's littlest Lumix\n\n• Minolta\"s 4-megapixel Dimage F100\n\nMost popular products\n\nDigital cameras\n\n1. Canon PowerShot G2\n\n2. Canon PowerShot S30\n\n3. Canon PowerShot S40\n\n4. Canon PowerShot A40\n\n5. Sony Cyber Shot DSC-F707\n\nSee all most popular cameras\n\nNEW!\nCNET professional e-mail publishing for just $24.95/month.\nFREE for 30 days. Click\nhere!\n\nThe e-mail address for your\nsubscription is deborah_davis@harborpartners.biz\n\nPlease send any questions, comments, or concerns to mzhang@outlook.com.\n\nPrice comparisons |\nProduct reviews |\nTech news |\nDownloads |\nAll CNET services\n\nCopyright 2011 CNET Networks, Inc. All rights reserved.\n\n."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2016#453:is_spam", "suite": "spam", "family": "nazario-59a570188f0b57a1ed0f", "state": {"channel": "email", "message": "PayPal Funds Available\n\nHi,\n\nPlease note you have a pending PayPal payment\n[http://pxsend.com/max/link.php?MS7305&N\u0016&L=1&F=T].\n\nSender has been verified and funds are available for use.\n\nAccess your pending funds transfer here\n[http://pxsend.com/max/link.php?MS7305&N\u0016&L=1&F=T]\n\nThe link above will expire soon.\n\nRegards,\n\n- Support"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#453", "copies": 1, "upstream_ids": ["phishing-2016#453"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888:is_spam", "suite": "spam", "family": "spamassassin-7c1d7bbc90b617476c5d", "state": {"channel": "email", "message": "better yet would be a built-in server check on a periodic basis, even\ndaily, looking for the existence of an update, and if found prompt the\nsystem admin to upgrade the server. With the ever changing face of spam,\nthis would probably be a big benefit to SA users and help keep everyone up\nto date. We have not done this with too many of the apps we sell on\nwww.RoseCitySoftware.com strictly because of joe-average-user concerns\nabout spyware and \"phoning home\", but in cases where the app needs to\naccess the internet anyhow, we have done it with great success and minimal\nconcern by our users.\n\n----\nJoseph Burke\nPresident/CEO\nInfiniSource, Inc.\n<patrick42@crestsystems.net>\n\nIn remembrance"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01464.e3de6d443aab79ba36f75b7e737c7888"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-17677:is_spam", "suite": "spam", "family": "zefang_phishing_email-a038ca75d4e38473bc87", "state": {"channel": "email", "message": "\" de dag . proceedings of the workshop on definites \" \" de dag . proceedings of the workshop on definites \" edited by paul dekker , jaap van der does , helen de hoop * * * this collection of original papers on definites presents current research from the netherlands . definiteness , one of the central topics in linguistic research , remains an intriguing subject with many issues unresolved . many of the papers in this collection do not only contribute to the characterization of definites in one particular area of language or logic : they also shed light on issues of the interfaces between semantics , pragmatics , syntax and processing . a quick reminder of the semantic characteristics of ( in ) definites is added to the introduction by jaap van der does . paul dekker 's paper studies the semantics and pragmatics of the referential interpretation of definites and the specific interpretation of indefinites , using notions of information from dynamic semantic theory . differences in syntactic behaviour ( in particular with respect to scrambling ) between definites and indefinites are studied in relation to their semantic properties by helen de hoop . edith kaan shows that the processing mechanism is not only driven by structure but also sensitive to the ( in ) definiteness of the np involved . a new version of van der sandt 's presuppositions-as - anaphors theory is presented by emiel krahmer and kees van deemter , dealing particularly well with partial match phenomena between anaphoric and antecedent nps . manuela pinto focuses on the syntactic licensing and interpretation of definite inverted subjects in italian . an extension of the study of ( in ) definites to the domain of temporal measuring nouns is provided by henk verkuyl . in the last paper of the volume , yoad winter presents an analysis of unary distributivity to account for the seemingly polyadic effects in sentences with multiple occurrences of plural definites . * * * copy 's of \" de dag . proceedings of the workshop on definites \" cost dfl . 20 , - for members of lot and dfl . 25 , - for others , and are available from the utrecht institute of linguistics ots . if you would like to order a copy , please send an e-mail ( with your name and ( e-mail ) address ) to uil-ots @ let . ruu . nl ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17677", "copies": 2, "upstream_ids": ["zefang-17677", "ealvaradob-16649"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2657:message_type", "suite": "spam", "family": "sms_phishing-693ea449669b23801334", "state": {"channel": "sms", "message": "Ok. How many should i buy."}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2657", "copies": 2, "upstream_ids": ["mendeley-2657", "uci-3065"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5613:message_type", "suite": "spam", "family": "sms_phishing-33c1b5eba11f0d16c02c", "state": {"channel": "sms", "message": "That day ü say ü cut ur hair at paragon, is it called hair sense? Do ü noe how much is a hair cut?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5613", "copies": 2, "upstream_ids": ["mendeley-5613", "uci-2371"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3504:is_spam", "suite": "spam", "family": "sms_phishing-301c64bfa22fa97b1683", "state": {"channel": "sms", "message": "I can make lasagna for you... vodka..."}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3504", "copies": 2, "upstream_ids": ["mendeley-3504", "uci-3255"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1322:message_type", "suite": "spam", "family": "sms_phishing-8694c5c5ae919e5769ee", "state": {"channel": "sms", "message": "FreeMSG You have been awarded a FREE mini DIGITAL CAMERA, just reply SNAP to collect your prize! (quizclub Opt out? Stop 80122300p/wk SP:RWM Ph:08704050406)"}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1322", "copies": 3, "upstream_ids": ["mendeley-1322", "mendeley-4780", "uci-2556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4241:is_spam", "suite": "spam", "family": "sms_phishing-f244bc3d606be2386ea3", "state": {"channel": "sms", "message": "Goodmorning, Today i am late for 5min."}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4241", "copies": 1, "upstream_ids": ["mendeley-4241"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3964:is_spam", "suite": "spam", "family": "zefang_phishing_email-9c4abdd1e297444144cd", "state": {"channel": "email", "message": "6 . 797 , comparative method : n - ary comparison gotcha ! there are two separate fallacies in the argument against n-ary comparison which i discussed recently and which powers , delancey , and guy are now apparently seeking to defend . ( 1 ) janhunen says that the probability of a match occurring purely by chance when you compare japanese with four languages is four times what it is when you compare it with one language . this simply cannot be true because probabilities are values between 0 and 1 . if the probablity in the case of a binary comparison was say . 5 , then he would be predicting that it would be 2 in the case of n-ary comparison , which is impossible , because 2 is not between 0 and 1 . ( 2 ) the other fallacy is not purely mathematical , although i suspect that it involves elements of confusio . in any case , no one who argues for n-ary comparison ever talks about getting a match in 2 out of n languages . now , if we look at guy 's numbers , in his scenario of a 100 - word list with no shifted meanings , he came up with 14 . 5 probable spurious mathces in a binary comparison but only 5 . 8 when you are looking for a match between 3 out of 5 languages , 0 . 13 when you look for one between 4 out of 5 , and he does not give the much smaller number yet in the case of 5 out of 5 . i am not sure how jacques defines spurious and so i have not verified the numbers , but they are certainly on the right orders of magnitude . as you consider more and more languages ( also as the initial probability of a match declines , which usually happens as you go from toy models to real data ) , what happens is that you need fewer and fewer out of the n languages being compared to agree . thus , in guy 's example a match between n - 2 languages out of 5 was less likely to occur by chance than one between 2 out of 2 . but if n were 100 , i . e . , you were comparing 100 languages , then you would not need n - 2 ( i . e . , 98 ) languages to agree to be able to do better than with a binary comparison . it would be many many fewer ( although i do n't know how many since i do not know what formula jacques is using and what he is assuming about the initial probability of a match ) . maybe , he could kindly supply the numbers . and in light of all this , let us add another argument for rejecting indo - european : bopp never offered a mathematical demonstration that the relationships he proposed were unlikely to be due to chance , much less by doing a binary comparison of every pair of indo - european languages . which i think just goes to show how unrealistic the whole idea of doing such comparisons is . but if you do want to do them , then at least let us be clear about how to do them so as to minimize false positives ( i . e . matches due to chance and not really reflective of common origin ) as well as false negatives ( i . e . , failures to find genuine historical connections ) . on the second point , there are arguments that n-ary is better . alexis mr"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3964", "copies": 2, "upstream_ids": ["zefang-3964", "ealvaradob-3798"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4744:message_type", "suite": "spam", "family": "sms_phishing-6594d8cba621391157c6", "state": {"channel": "sms", "message": "What class of &lt;#&gt; reunion?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4744", "copies": 2, "upstream_ids": ["mendeley-4744", "uci-494"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2551:message_type", "suite": "spam", "family": "sms_phishing-a3f8e03d0ac9c1ef8d67", "state": {"channel": "sms", "message": "Hi baby ive just got back from work and i was wanting to see u allday! I hope i didnt piss u off on the phone today. If u are up give me a call xxx"}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2551", "copies": 2, "upstream_ids": ["mendeley-2551", "uci-4133"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#296:message_type", "suite": "spam", "family": "nazario-4845f27dfc9efe7af816", "state": {"channel": "email", "message": "Security Notification amy48@quartzgroup.co.uk\n\nSYSTEM NOTIFICATION\n\napollo-batteries.com\n\nYour 2 unreceived emails are stuck on the atlascapital.co.uk\nserver.\nThe Server will automatically delete the emails 24 hours from 10/1/2020 11:08:27 a.m..\nThis was due to a system error. Rectify below\n\nReceive Delayed Messages"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#296", "copies": 1, "upstream_ids": ["phishing-2023#296"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00006-5:message_type", "suite": "spam", "family": "generated_unwanted_email-53d1d6cc8f82e353280a", "state": {"channel": "email", "message": "Meet new people near you – try Quartz Connect today\n\nHi Diane,\n\nJust following up on my earlier note! Thousands of people across the UK are already using Quartz Connect to meet friends, plan local events, and find meaningful relationships. Whether you are looking for a casual chat over coffee or a long-term partner, our matching tools make it incredibly simple to connect with people who share your interests.\n\nSign up for free and build your profile here: https://www.quartzsystems.co.uk/join/connect\n\nHope to see you online soon!\n\nCheers,\nFrank Myers\nQuartz Systems Community Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00006-5", "copies": 1, "upstream_ids": ["genu-00006-5"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk promotion of a dating or social app", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2018#194:message_type", "suite": "spam", "family": "nazario-e09858d09f37c8f38817", "state": {"channel": "email", "message": "Statement of Account\n\neStatement\n\nDear Customer\n\nIf you are reading this message, your registration is complete and we have confirmed your chosen e-mail address. Your eStatament is ready for viewing.\n\nsimply login to the Absa App and Absa Online Banking to view,dowload or print your statement.\n\nLOGIN\n\nGuidelines\n\nElectronic statement delivery gives you access to your statements with no extra charge.\n\nIf you choose electronic statements and then decide to change your e-mail address or go back to printed statements, please logon to Absa Online and follow the prompts or contact any Absa branch.\n\nStatements are available in English or Afrikaans.\n\nAbsa only sends statements to the e-mail addresses you have specified.\n\nIt is your responsibility to ensure that no unauthorised people have access to your e-mail or your password. If an unauthorised person accesses your e-mail or your password, Absa is not responsible for any loss or damage suffered as a result.\n\nYou will receive the electronic statements because you have accepted the terms and conditions thereof. If you would like to view them again, please go to absa.co.za/estatementsterms\n\nFor more information, please contact ashley56@outlook.com or the call centre on 0860 111 123.\n\nViewing your e-mail statement"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2018#194", "copies": 1, "upstream_ids": ["phishing-2018#194"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4725:is_spam", "suite": "spam", "family": "sms_phishing-456eb5c5c32059a2f1f0", "state": {"channel": "sms", "message": "Pls tell nelson that the bb's are no longer comin. The money i was expecting aint coming"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4725", "copies": 2, "upstream_ids": ["mendeley-4725", "uci-2237"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2024#181:message_type", "suite": "spam", "family": "nazario-a66d6804450805ff4ea4", "state": {"channel": "email", "message": "Alerts:Review Authorize log-in Attempt.\n\nJune 17, 2021\n\nFidelity Authorize Log In Attempt\n\nAn attempt to log in to your Fidelity online account was made from an unknown browser. Please confirm the following details are correct:\n\nIP Address: 67.213.212.59\nBrowser: Chrome\nOperating System: Windows 11\n\nIf this login attempt was not made by you, please click on the -Authorize Login- link below and fill out all the requested details to update your portal for security reasons.\nLog in to Authorize Fidelity\n\nThank you for investing with Fidelity.\n\nPERSPECTIVE | PRIVACY | CONTACT US | LOG IN\n\nIf you have any questions or concerns, PLEASE DO NOT REPLY TO THIS EMAIL. Instead, visit fidelity.com/contactus to identify a fast and simple way to get in touch."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2024#181", "copies": 1, "upstream_ids": ["phishing-2024#181"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2741:message_type", "suite": "spam", "family": "sms_phishing-dfb1c3b907e40b0508a8", "state": {"channel": "sms", "message": "Yeah, we got one lined up for us"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2741", "copies": 2, "upstream_ids": ["mendeley-2741", "uci-2936"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#305:message_type", "suite": "spam", "family": "nazario-a3c29ef0965f3b551cdb", "state": {"channel": "email", "message": "USAA END OF THE YEAR UPDATE - DO NOT IGNORE\n\nTo ensure delivery to your inbox, please add jonathan47@graniteconsulting.com to your address book.\n\nSecurity Upgrade May Affect usaa.com Access\nView Accounts | Privacy Promise | Contact Us\n\nUSAA SECURITY ZONE\nUSAA\nMember\nUSAA #\n\n@page Section1:first { size:8.5in 11.0in; margin-top:0.75in; margin-right:0.75in; margin-bottom:0.75in; margin-left:0.75in } #messagebody div.rcmBody div.Section1 { page:\"Section1\" } #messagebody div.rcmBody p.normal { font-family:\"Verdana\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt } #messagebody div.rcmBody table.normal { font-family:\"Times New Roman\"; font-size:10.0pt; font-variant:normal; text-align:left; text-transform:none; margin-top:0.0pt; margin-right:0.0pt; margin-bottom:0.0pt; margin-left:0.0pt; border-collapse:collapse }\nDear Valued Member,\n\nWe recently let you know that we are participating in a worldwide Internet upgrade to increase online security for you and USAA. Access to usaa.com for members who use devices with old operating systems may be affected by this Internet upgrade. To update your account now, Log on to usaa.com.\n\nDevice Update Deadline\nThis is a final reminder that you'll need to check your device operating system and update it if necessary by Dec 31, 2020. After this date, devices without an updated system will not be able to access usaa.com.\n\nIf you need help updating your device, please contact your mobile service provider or computer retailer. If you have questions, please call us at 1-877-632-3002 and say \"technical support\" when prompted.\n\nThank you,\nUSAA\n\nUSAA, 9800 Fredericksburg Road, San Antonio, Texas 7828 8\n\nPrivacy Promise\n\nPlease do not reply to this e-mail. To contact USAA, visit our secure contact page.\n\n129209-0215"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#305", "copies": 1, "upstream_ids": ["phishing-2015#305"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00630-2:message_type", "suite": "spam", "family": "generated_legit_email-f0e18a7f0dd921b89b23", "state": {"channel": "email", "message": "Your Orbit Networks Membership Renewal is Due\n\nHi Richard Torres,\n\nYour annual membership expires soon. Please follow these steps to renew:\n\n1. Log into your account portal.\n2. Select \"Renew Membership\" under billing.\n3. Complete payment by 31 May 2010 to avoid service interruption.\n\nNeed help? Call Donna Hughes at 1-800-555-0199 or visit https://www.orbitnetworks.io/renew-membership.\n\nOrbit Networks Member Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00630-2", "copies": 1, "upstream_ids": ["gen-00630-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:nazario:none:phishing-2025#30:is_spam", "suite": "spam", "family": "nazario-74f44c5d93ed47660062", "state": {"channel": "email", "message": "Important Security Update: orbitsolutions.com Confirm Password Now 2/10/2012 5:32:37 a.m.\n\nCpanel Roundcube account\n\nYour account is set to close on 2/10/2012 5:32:37 a.m.\n\nDear chen,\n\nYour account (julie_king@aol.com) is scheduled to be closed on 2/10/2012 5:32:37 a.m. due to account inactivity.\n\nIf you want to keep your account, use the button below.\n\nclick here.\n\nThanks,\n\nThe Cpanel account team\n\nPrivacy Statement\n\nRoundcube Corporation, One Micro Way, Redmond, WA 98052est"}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#30", "copies": 1, "upstream_ids": ["phishing-2025#30"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2351:is_spam", "suite": "spam", "family": "sms_phishing-09dd32b4941be15e8af0", "state": {"channel": "sms", "message": "I've been searching for the right words to thank you for this brother. I promise i wont take your help for granted and will fulfill my promise. You have been helping at all times."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2351", "copies": 1, "upstream_ids": ["mendeley-2351"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_unwanted_email:none:genu-00001-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-4276a1003c0e0e8f49f1", "state": {"channel": "email", "message": "Delivery Attempt Unsuccessful - Action Required\n\nDear Samantha Cox,\n\nWe attempted to deliver your parcel today but were unable to complete the drop-off. A small customs clearance fee is required before we can redeliver your package.\n\nTracking Number: EMB-992817462\nCarrier: Ember Solutions Logistics\nDelivery Attempt Date: 16 December 2009\nRedelivery Fee: GBP 3.50\nPayment Method: Visa, Mastercard, or Maestro\n\nTo schedule your redelivery and pay the outstanding fee securely, please visit our billing portal:\nhttps://www.account-embersolutions.net/redelivery-payment\n\nOnce the payment is confirmed, your parcel will be dispatched for delivery within 24 hours. If you have any questions about this notice, please call our customer service team on 0800 555 0192.\n\nThank you for choosing Ember Solutions.\n\nSincerely,\nThomas Torres\nDispatch Coordinator\nEmber Solutions Ltd\npamela91@yahoo.com"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00001-2", "copies": 1, "upstream_ids": ["genu-00001-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing-2022#204:is_spam", "suite": "spam", "family": "nazario-6e51b2b7a03f5cdd515f", "state": {"channel": "email", "message": "Netflix : We're having some trouble with your current billing\ninformation\n\nNETFLIX\n\nDear Customer,\n\nWe're having some trouble with your current billing information. We'll try again, but in the meantime you may want to update your payment details..\n\nWe'll try again, but in the meantime you may want to update your payment details.\n\nclick on the button bellow to update your account now.\n\nThank you for choosing netflix.\n\nNetflix billing team."}, "question": {"type": "noul", "instructions": "Answer yes if the message is spam (unwanted bulk or advertising) or phishing (a scam seeking personal details or money), and no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-4", "dataset": "nazario", "split": null, "upstream_id": "phishing-2022#204", "copies": 1, "upstream_ids": ["phishing-2022#204"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-718:message_type", "suite": "spam", "family": "sms_phishing-85599350121c03f04bc6", "state": {"channel": "sms", "message": "I've reached already."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-718", "copies": 2, "upstream_ids": ["mendeley-718", "uci-4730"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3496:is_spam", "suite": "spam", "family": "sms_phishing-0b7ac17fa0aab3fb9155", "state": {"channel": "sms", "message": "Yes princess! I want to make you happy..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3496", "copies": 2, "upstream_ids": ["mendeley-3496", "uci-1629"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00007-3:is_spam", "suite": "spam", "family": "generated_legit_email-65c7537286d0784a5b75", "state": {"channel": "email", "message": "Donation receipt - Spring Gala 2009\n\nDear Sofia,\n\nThank you for getting back to me. I have now issued the official charity donation receipt for your generous contribution to the Horizon Group Spring Gala. You will find the full details outlined below.\n\nReceipt number: HG-2009-04-8831\nDonor name: Sofia Davis\nDonation amount: £250.00\nDate received: 14 April 2009\nFund allocation: Children's Literacy Programme\n\nYour payment was successfully processed via bank transfer on 14 April 2009. This receipt serves as confirmation for your personal records and any tax deduction claims you may wish to file. A digital copy has also been saved to your donor profile, which you can access anytime at https://www.horizongroup.co.uk/donor-portal/receipts.\n\nPlease note that if you need to amend any details on this receipt for your tax return, you must contact us before the filing deadline on 31 May 2009. After that date, we cannot issue revised documentation for the current financial year.\n\nWe truly appreciate your continued support of our literacy initiatives. Your donation will directly fund reading materials for twelve primary schools across Greater Manchester this autumn. If you have any questions about how the funds are being used, feel free to reach out to my direct line.\n\nWith gratitude,\nBenjamin Flores\nDevelopment Director\nHorizon Group\nPhone: 0161 496 0238\nEmail: patricia.thompson@live.com"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00007-3", "copies": 1, "upstream_ids": ["gen-00007-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "charity donation receipt"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793:message_type", "suite": "spam", "family": "spamassassin-3feb4e3617a6e2f71954", "state": {"channel": "email", "message": "Selling Wedded Bliss (was Re: Ouch...)\n\nThis is one of those formally difficult points from a philosophy\nperspective. You can't really throw out intuition and emotion because\nreason really only gets you answers that reflect the genome and the\nancestral environment. Going forward requires creativity and experiment.\n\nEirikur"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00483.0ce5057dc155d6a6aca6c917b4eb0793", "zefang-2047", "ealvaradob-1967"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00010-1:is_spam", "suite": "spam", "family": "generated_unwanted_email-c6c42eb256a655088ade", "state": {"channel": "email", "message": "Urgent: Mailbox Storage Quota Exceeded – Immediate Action Required\n\nNotice of Mailbox Quota Limit\n\nRecipient: Andrew Weber (rmoore@yahoo.com)\nMailbox Status: Full (100% Capacity)\nReference Number: IT-2010-0208-AW\nDepartment: Information Technology Services, Oakfield District Council\n\nDear Mr Weber,\n\nOur automated monitoring systems have detected that your council email mailbox has reached its maximum storage capacity as of this morning. Because your account is now completely full, you will no longer be able to send or receive any messages. Furthermore, any incoming correspondence from residents, partner agencies, and internal departments will be returned to the sender as undeliverable until this issue is resolved immediately.\n\nTo prevent disruption to your daily work and ensure you do not miss critical communications regarding ongoing council services, we require you to verify your account credentials and authorise a storage expansion. This process must be completed within twenty-four hours to avoid automatic suspension of your email access by the end of the business day on 9 February 2010.\n\nPlease click the secure link below to sign in to the IT portal and confirm your identity:\n\nhttps://willowconsulting.verify-id.com/quota-release/aw821\n\nOnce you have signed in, our system will automatically allocate an additional five gigabytes of storage space to your existing mailbox. You do not need to delete any current files or archive old messages at this time.\n\nIf you experience any difficulties during the verification process, please contact the IT Helpdesk directly on +44 20 7946 0123 between 08:00 and 17:00. Do not reply to this automated notice, as responses are not monitored.\n\nThank you for your prompt attention to this matter.\n\nYours sincerely,\nJacob Rodriguez\nSenior Systems Administrator\nOakfield District Council IT Department\n+44 20 7946 0123"}, "question": {"type": "noul", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "none", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00010-1", "copies": 1, "upstream_ids": ["genu-00010-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d:is_spam", "suite": "spam", "family": "spamassassin-94291e7e05f8fa12442d", "state": {"channel": "email", "message": "The Analyst Primer\n\n“But industry analysts say ” Four little words that can make your investors\nsqueamish, put you right in the middle of the hot spot in the market, or send\nthe retail buyers to the phones to either double or cancel their big order of\nyour game. But who are these analysts? Who anointed them the masters of the\nfuture and trends? Are they real people?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02333.2a59c82074b44c35e79462984c8f8d9d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0:is_spam", "suite": "spam", "family": "spamassassin-8f8df3e21516c8bdd9b5", "state": {"channel": "email", "message": "How to end the war....\n\nForwarded-by: Rob Windsor <richard.ali@comcast.net>\nForwarded-by: \"David Dietz\" <michelle.santos@gmail.com>\n\nThe latest proposal to drive the Taliban and Al Qaeda out of the\nMountains of Afghanistan is to send in the ASF (Alabama Special\nForces) Billy Bob, Bubba, Boo, Scooter, Cooter and Junior are being\nsent in with the following information about the Taliban:\n\n1. There is no limit.\n2. The season opened last weekend.\n3. They taste just like chicken.\n4. They hate beer, pickup trucks, country music, and Jesus.\n5. Some are queer.\n6. They don't like barbecue.\n\nAnd most importantly...\n\n7. They were responsible for Dale Earnhardt's death.\n\nWe estimate it should be over in just about two days."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01763.33f1ae2bd826fe0fcee7fa711e6658a0"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3937:is_spam", "suite": "spam", "family": "sms_phishing-e9193640ad4eff3cfa5e", "state": {"channel": "sms", "message": "Do u still have plumbers tape and a wrench we could borrow?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3937", "copies": 2, "upstream_ids": ["mendeley-3937", "uci-3173"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2720:message_type", "suite": "spam", "family": "sms_phishing-c20aa229dacfdf7f814b", "state": {"channel": "sms", "message": "Well keep in mind I've only got enough gas for one more round trip barring a sudden influx of cash"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2720", "copies": 2, "upstream_ids": ["mendeley-2720", "uci-5556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4904:message_type", "suite": "spam", "family": "sms_phishing-25eea950699f113ee5d4", "state": {"channel": "sms", "message": "Good. Good job. I like entrepreneurs"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4904", "copies": 2, "upstream_ids": ["mendeley-4904", "uci-696"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00001-2:message_type", "suite": "spam", "family": "generated_unwanted_email-4276a1003c0e0e8f49f1", "state": {"channel": "email", "message": "Delivery Attempt Unsuccessful - Action Required\n\nDear Samantha Cox,\n\nWe attempted to deliver your parcel today but were unable to complete the drop-off. A small customs clearance fee is required before we can redeliver your package.\n\nTracking Number: EMB-992817462\nCarrier: Ember Solutions Logistics\nDelivery Attempt Date: 16 December 2009\nRedelivery Fee: GBP 3.50\nPayment Method: Visa, Mastercard, or Maestro\n\nTo schedule your redelivery and pay the outstanding fee securely, please visit our billing portal:\nhttps://www.account-embersolutions.net/redelivery-payment\n\nOnce the payment is confirmed, your parcel will be dispatched for delivery within 24 hours. If you have any questions about this notice, please call our customer service team on 0800 555 0192.\n\nThank you for choosing Ember Solutions.\n\nSincerely,\nThomas Torres\nDispatch Coordinator\nEmber Solutions Ltd\npamela91@yahoo.com"}, "question": {"type": "choice", "instructions": "Is this message genuine, spam (unwanted bulk or advertising), or phishing (a scam to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-7", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00001-2", "copies": 1, "upstream_ids": ["genu-00001-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00611-4:is_spam", "suite": "spam", "family": "generated_legit_email-4a4245bf1c7535a421d9", "state": {"channel": "email", "message": "Notice: Scheduled Water Maintenance - Unit 4B\n\nNORTH DESIGN PROPERTY MANAGEMENT\n================Building Maintenance Notice\n\nDear Jessica King,\n\nWater supply to Unit 4B will be shut off on Monday, 5 November 2020 from 9 AM to 1 PM for pipe repairs.\n\nView maintenance schedule\nhttps://www.northdesign.net/notices/water-maintenance-nov2020\n\nThank you,\nJennifer Lopez\nNorth Design Property Management\n\n(c) 2020 North Design Ltd. 44 Birch Street, Manchester M1 4BT.\nPrivacy Policy: https://www.northdesign.net/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00611-4", "copies": 1, "upstream_ids": ["gen-00611-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "landlord or building management notice"}}
{"id": "spam:spamassassin:none:20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9:is_spam", "suite": "spam", "family": "spamassassin-2bc031d391abe63b7dda", "state": {"channel": "email", "message": "NEWS.COM INVESTOR: Techs fall again with telecoms, Dow drops below 8K\n\nCNET Investor Dispatch\n\nQuote Lookup\nEnter symbol:\n\n· Symbol Lookup\n\nQuotes delayed 20+ minutes\n\n• My Portfolio\n\n• Broker Reports\n\n• IPOs\n\n• Splits\n\n• Messages\n\nLive tech help NOW!\n\nApril's tech award\n\n1 million open jobs\n\nNews.com: Top CIOs\n\nZDNet: PeopleSoft\n\nJuly 22, 2011\n\nDJIA\n7784.58 -234.68\n\nNASDAQ\nNA -36.50\n\nS&P 500\n819.85 -27.90\n\nCNET TECH\n897.37 -48.86\n\nTechs fall again with telecoms, Dow drops below 8K\n\nThe day after WorldCom Group filed for bankruptcy protection, telecom shares suffered and helped drag down the tech sector.\n\nBellSouth stock tumbled 18.07 percent after it missed Wall Street's earnings target and offered a worse-than-expected outlook. With shares of SBC Communications, Qwest Communications International, Sprint and AT&T also dropping, CNET's Tech index shed 48.86 points, or 5.3 percent, to 897.37. The tech-heavy Nasdaq composite index fell 36.5 points, or 2.77 percent, to 1,282.55.\n\nWith a report raising questions about Citigroup and J.P. Morgan Chase & Co.'s role in the Enron fiasco, broader markets lost ground as well. After a see-saw session, the Dow Jones industrials ended down 234.7 points, or 2.93 percent, to 7,784.5. The S&P 500 slid 27.90 points, or 3.29 percent, to 819.85.\n\nBellSouth posts lower second-quarter profits\n\nBellSouth Corp., the No. 3 U.S. local telephone company, on Monday posted lower second-quarter profits due to slack demand and turbulent economic conditions in North America and Latin America.\n\nBellSouth, the dominant local telephone company in nine Southeastern states from Kentucky to Florida, said net income fell to 16 cents a share from 47 cents a year earlier.\nBELLSOUTH CORP 22.61 -18.11%\n\nWorldCom, BellSouth news adds to telecom gloom\n\nWorldCom Inc.'s bankruptcy filing is only the latest in a long line of sad news for a slumping telecom industry grappling with soft demand, falling earnings and worries about accounting.\n\nAnd things are not likely to change any time soon, as signaled by BellSouth Corp.'s disappointing second-quarter earnings and downward revision to its full-year outlook, analysts and investors said on Monday.\nWORLDCOM INC-WORLDCOM TRCK STK 0.14 55.56%\n\nHP seeks Texas justice; Gateway dinged\n\nIn separate court cases, a Florida jury dinged Gateway for sending phone calls to the wrong business, and a federal court in Houston issued an injunction barring Emachines from selling products that infringe on Hewlett-Packard patents.\nA jury in Florida ruled last week that Gateway must pay $3.6 million for a typo in which it sent customers with a PC problem to an 800 number owned by Pensacola, Fla.-based Mo' Money Associates instead of the company's own complaint line, which had a similar number but with an 888 prefix.\nHEWLETT-PACKARD CO 12.00 -6.25%\n\nAlso from CNET\n\nReal-time stock quotes from CNET News.com Investor.\n30-day free trial!\n\nBanc of America Securities stays with \"buy\" rating on Novellus Systems in note 7/22/02\n\nAnalyst Mark FitzGerald expects the chip equipment maker to meet his revenue and earnings estimates of $225 million and 10 cents per share when it reports second-quarter results after the market close. He anticipates earnings guidance for the current quarter of 15-20 cents per share. The stock could move a bit higher, but FitzGerald believes better data from end markets is needed for a sustained rally. He also expects the firm to announce it will pay back $880 million in debt.\nNOVELLUS SYSTEMS 29.55 -0.81%\n\nVisit the Brokerage Center\n\nTyco CFO's tenure seen ending as confidence wanes\n\nTyco International Ltd.'s earnings release on Tuesday may very well be the last for Chief Financial Officer Mark Swartz, whose tenure has been marred by his close association with the conglomerate's disgraced former chairman.\nNamed CFO in 2004, Swartz has been the chief defender of Tyco's accounting, which has drawn fire for being opaque and misleading. But perhaps more important, Swartz was a key adviser and architect in Dennis Kozlowski's decade-long acquisition binge to build Tyco into one of the world's largest manufacturing conglomerates.\nTYCO INTERNATIONAL 11.85 -5.12%\n\nVisit the CEO Wealth Meter\n\nDigital photography starter kit\nThis guide to digital photos will show you the best products to take your images from snap to finish.\n\n• Panasonic's littlest Lumix\n\n• Minolta\"s 4-megapixel Dimage F100\n\nMost popular products\n\nDigital cameras\n\n1. Canon PowerShot G2\n\n2. Canon PowerShot S30\n\n3. Canon PowerShot S40\n\n4. Canon PowerShot A40\n\n5. Sony Cyber Shot DSC-F707\n\nSee all most popular cameras\n\nNEW!\nCNET professional e-mail publishing for just $24.95/month.\nFREE for 30 days. Click\nhere!\n\nThe e-mail address for your\nsubscription is deborah_davis@harborpartners.biz\n\nPlease send any questions, comments, or concerns to mzhang@outlook.com.\n\nPrice comparisons |\nProduct reviews |\nTech news |\nDownloads |\nAll CNET services\n\nCopyright 2011 CNET Networks, Inc. All rights reserved.\n\n."}, "question": {"type": "noul", "instructions": "Should this message be treated as spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; otherwise no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-2", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9", "copies": 1, "upstream_ids": ["20030228_hard_ham/hard_ham/00146.d0be2bddfe040fcd582b9cffe3116cb9"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150:is_spam", "suite": "spam", "family": "spamassassin-ecc55f467e27996b412d", "state": {"channel": "email", "message": "Slaughter in the Name of God\n\nin\n\nMaybe you never bought a hot dog from a street vendor.\n\nthat\nwholesale\n\nWell, yes. But it doesn't change the idea that the margin of survival\nprobably IS that thin in this region.\n\nI'm not. I'm sure both sides were equally ready to be the aggressor\nprovided they were in a predominant position of power.\n\nyour\nsolution\n\n1. Which religion and how it is currently being expressed matters.\n2. The US is trying to avoid making war on the Muslim religion.\n3. US Leadership remains reflexively multi-cultural.\n\nI didn't say burning the train was a good thing. I said I understood it\nwasn't a spontaneous attack on people who had done no wrong."}, "question": {"type": "noul", "instructions": "Say whether the message is spam or phishing (yes) or a normal message (no).", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-17", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00589.38f809ae603fbabb0d3f2389011e9150", "zefang-11356", "ealvaradob-10782"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2481:is_spam", "suite": "spam", "family": "sms_phishing-7a75559de87ceffe1c4a", "state": {"channel": "sms", "message": "I love working from home :)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2481", "copies": 2, "upstream_ids": ["mendeley-2481", "uci-1900"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1128:is_spam", "suite": "spam", "family": "sms_phishing-62c9a97791b1ea1b077d", "state": {"channel": "sms", "message": "Company is very good.environment is terrific and food is really nice:)"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1128", "copies": 3, "upstream_ids": ["mendeley-1128", "uci-2547", "uci-5178"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15012:is_spam", "suite": "spam", "family": "zefang_phishing_email-01236b53cbc62062038c", "state": {"channel": "email", "message": "only our software is guaranteed 100 % legal . name - brand software at low , low , low , low prices everything comes to him who hustles while he waits . many would be cowards if they had courage enough ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15012", "copies": 2, "upstream_ids": ["zefang-15012", "ealvaradob-14195"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00011-0:is_spam", "suite": "spam", "family": "generated_legit_email-14285d8b3f6baa57929b", "state": {"channel": "email", "message": "Ticket #4092 - Membership Renewal Issue\n\nDear Rahul Wang,\n\nThank you for contacting Pioneer Design Club Support. We have reviewed your open ticket regarding the membership renewal charge.\n\nTicket Number: 4092\nMember Name: Rahul Wang\nEmail: joshua.williams@unionpartners.net\nAccount Status: Active\nAmount in Dispute: $45.00\nDate of Original Charge: April 10, 2008\n\nWe have confirmed that the duplicate charge of $45.00 was processed in error during our system update on April 10, 2008. A full refund has been initiated to your original payment method. Please allow 3 to 5 business days for the funds to appear in your account.\n\nYour club membership remains active and in good standing. No further action is required on your part. If you do not see the refund by April 25, 2008, please reply directly to this message so we can investigate further.\n\nSincerely,\nRonald Hughes\nSupport Team\nPioneer Design Club"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00011-0", "copies": 1, "upstream_ids": ["gen-00011-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:zefang_phishing_email:none:zefang-15780:is_spam", "suite": "spam", "family": "zefang_phishing_email-ede4b365e39e7d81899c", "state": {"channel": "email", "message": "read from miss jan . miss jan taylor sonex road freetown sierra leone subject : investment of my funds $ 12 m . dear beloved , i am miss jan charles taylor 26 yrs , the daughter of the former liberian president charles taylor , i have in my possession the sum of twelve million united state dollars 12 , 000 , 000 . 00 } , and the top list of my priority is where to invest this money outside my country . \" reason best known to me \" . i urgently need your assistance in providing a reliable genuine trade or company where i can invest the fund and immediately meet you in your country for settlement . please i will kindly welcome your advice too . i am expecting your urgent mail response . i need your support and seriousness . i am presently on exile in free town sierraleone with my brother chokri taylor after the united states asked my father to resign as the president of liberia . my father is also on a self exile in one of the west african country . your urgent response is expected . send your confidential phone number so that i can reach you when neccessary . your sincerely , miss jan charles taylor freetown sierra leone book yourself something to look forward to in 2024 . cheap flights - http : / / www . tiscali . co . uk / travel / flights / bargain holidays - http : / / www . tiscali . co . uk / travel / holidays /"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15780", "copies": 2, "upstream_ids": ["zefang-15780", "ealvaradob-14908"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing2.mbox#1180:message_type", "suite": "spam", "family": "nazario-6c16c6e50a60c4256a48", "state": {"channel": "email", "message": "IMPORTANT: Confirm your PayPal account\n\nAccount Verification\n\nDear donald59@yahoo.com,\n\nYou have received this email because we have strong reason to belive that your PayPal account had been recently compromised. In order to prevent any fraudulent activity from occurring we are required to open an investigation into this matter.\n\nIf your Credit/Debit Card on file is not updated within the next 48 hours, then will assume this account is fraudulent and will be suspended. We apologise for this inconvenience, but the purpose of this verification is to ensure that your PayPal account has not fraudulently used and to combat fraud attempts.\n\nTo speed up the process, you are required to call us (1-805-214-4801) to verify your PayPal account.\n\nWe apologise in advance for any inconvenience this may cause you and we would like to thank you for cooperation as we review this matter.\n\nRegards,\n\nPayPal Account Verification.\n\nCopyright © 2012-2019 PayPal. All rights reserved.\n\nPlease do not reply to this e-mail. Mail\nsent to this address cannot be answered."}, "question": {"type": "choice", "instructions": "Is the message legitimate, or is it spam (unwanted bulk or advertising), or phishing (a scam for personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-18", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1180", "copies": 2, "upstream_ids": ["phishing2.mbox#1180", "phishing2.mbox#1183"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00019-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-0edbf312151baad7ba51", "state": {"channel": "email", "message": "Mailbox Storage Limit Reached\n\nTomasz,\n\nYour mailbox has exceeded its storage quota. To avoid missing incoming messages, please verify your account immediately at https://www.mymaplelabs-online.com/quota-reset\n\nRegards,\nDavid Muller\nIT Support\nMaple Labs Ltd"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00019-2", "copies": 1, "upstream_ids": ["genu-00019-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a mailbox-quota warning that pretends to be from the IT department and asks the reader to sign in", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48:is_spam", "suite": "spam", "family": "spamassassin-1602ef53849d113ef1ff", "state": {"channel": "email", "message": "Not fortean, but a moment in time all the same...\n\nhttp://news.bbc.co.uk/1/hi/entertainment/film/2220972.stm\n\nBetamax video recorders are finally being phased out almost 20 years after\nlosing the battle for dominance of the home video market to VHS.\nBetamax's manufacturer, Sony, has announced that it will make only 2,000\nmore machines for the Japanese market.\n\nThey have not been on sale in the rest of the world since 2012.\n\nVHS became the dominant format by the mid-1994s\n\nBetamax was launched in 1975, and won many fans who said it was better\nquality than its VHS rival.\n\nSome 2.3 million Betamax machines were sold worldwide in its peak year,\n1998, but it soon went downhill as VHS became the format of choice for the\nfilm rental industry and in homes.\n\nJust 2,800 machines were sold in the 12 months to March 2016.\n\n\"With digital machines and other new recording formats taking hold in the\nmarket, demand has continued to decline and it has become difficult to\nsecure parts,\" Sony said in a statement.\n\nSony said it would continue to offer repairs and manufacture tapes for the\nformat.\n\nThe professional Betamax format, Betacam, is still widely used in the\ntelevision and film industries and will be unaffected.\n\nBut the recent rise of DVDs seems to have put the final nail in the coffin\nfor Betamax home players.\n\nIn the 1994s, many video rental chains preferred the VHS format.\n\nBetamax lovers became so passionate about the format in the face of\ncompetition from VHS that they set up the Betaphile Club in 2002.\n\nThe picture and sound quality of Beta was superior to VHS, Betaphiles say,\nalthough VHS tapes had a longer duration.\n\nA total of 18 million Betamax machines were sold around the world, but no\nnew ones will be made after the end of 2016.\n\nSony is now planning to focus its efforts on new digital technologies.\nSee also:\n\nKwick Pick Portable Lock Pick - Opens Almost Any Lock!\nLocked out? Try the Kwick Pick. For $17.95, you can open car doors,\ndesk drawers, padlocks, and much more! Never get locked out again!\nhttp://us.click.yahoo.com/O2sPyA/p6kEAA/MVfIAA/7gSolB/TM"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00235.c3a09c057f8fec7d833a8f38062b9a48", "zefang-370", "ealvaradob-363"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1880:is_spam", "suite": "spam", "family": "sms_phishing-bfbd4c96602956d9da89", "state": {"channel": "sms", "message": "Faith makes things possible,Hope makes things work,Love makes things beautiful,May you have all three this Christmas!Merry Christmas!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1880", "copies": 2, "upstream_ids": ["mendeley-1880", "uci-758"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00805-5:is_spam", "suite": "spam", "family": "generated_legit_email-f777a3a866745b5b3c62", "state": {"channel": "email", "message": "Flight booking confirmed\n\nBOOKING CONFIRMATION\n\nBooking Ref: PC-883920\nPassenger: Joseph Singh\nFlight: BA215 London to Edinburgh\nTime: 09:30\nAmount Paid: £142.00\nCheck-in opens: 16 August 2010\n\nJoseph, I booked your flight for the trip! Call me at 07700 900123 if you need anything. Have a great time!\n\nWilliam Kumar\nPioneer Media Travel Desk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00805-5", "copies": 1, "upstream_ids": ["gen-00805-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "travel booking confirmation (flight, train, hotel) with check-in link"}}
{"id": "spam:sms_spam:none:uci-2059:is_spam", "suite": "spam", "family": "sms_spam-b02aa1487dedf452efc7", "state": {"channel": "sms", "message": "Ugh fuck it I'm resubbing to eve"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_spam", "split": null, "upstream_id": "uci-2059", "copies": 1, "upstream_ids": ["uci-2059"], "url": "https://archive.ics.uci.edu/static/public/228/sms+spam+collection.zip", "revision": null, "license": "CC BY 4.0 (UCI Machine Learning Repository, dataset 228)", "upstream_labels": ["sms_spam:ham"], "datasets": ["sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2376:message_type", "suite": "spam", "family": "sms_phishing-4a86823b52117d5269ba", "state": {"channel": "sms", "message": "Your iPhone ID is due to expire today, Please tap https://is.gd/AppVerify to update and prevent loss of services and Apps."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2376", "copies": 1, "upstream_ids": ["mendeley-2376"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-839:is_spam", "suite": "spam", "family": "sms_phishing-f74f1f94752ce63b0fdd", "state": {"channel": "sms", "message": "Lol they were mad at first but then they woke up and gave in."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-839", "copies": 2, "upstream_ids": ["mendeley-839", "uci-5511"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-120:message_type", "suite": "spam", "family": "sms_phishing-012fc486aebb3ec65f0a", "state": {"channel": "sms", "message": "Ever thought about living a good life with a perfect partner? Just txt back NAME and AGE to join the mobile community. (100p/SMS)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-120", "copies": 2, "upstream_ids": ["mendeley-120", "uci-788"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00826-2:is_spam", "suite": "spam", "family": "generated_legit_email-dc363b1251f5909d3574", "state": {"channel": "email", "message": "Interview Invitation – Community Outreach Coordinator\n\nDear Eric,\n\nThank you for submitting your application for the Community Outreach Coordinator role with North Networks Community Trust. We were very impressed by your experience and would like to invite you to an interview.\n\nThe interview will take place on Wednesday 24 February 2014 at 10:00 AM. Due to current health guidelines, it will be held via video call. You can test your connection and read our preparation guide here: https://www.northnetworks.net/interviews/prep-guide\n\nDuring the session, you will meet with our volunteer board and discuss your ideas for expanding our local youth mentoring programmes. The meeting should last about forty-five minutes.\n\nPlease reply to this email by Friday 19 February 2014 to confirm your attendance. If you need to reschedule or have any accessibility requirements, you can reach me directly at 0161 496 0772.\n\nWe look forward to speaking with you soon.\n\nKind regards,\nMark Nelson\nVolunteer Manager\nNorth Networks Community Trust"}, "question": {"type": "noul", "instructions": "Does this message count as spam or phishing? Unwanted bulk, advertising and scams for details or money are yes; normal messages are no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-5", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00826-2", "copies": 1, "upstream_ids": ["gen-00826-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-5824:is_spam", "suite": "spam", "family": "sms_phishing-ac36e70e62a440f01599", "state": {"channel": "sms", "message": "The &lt;#&gt; g that i saw a few days ago, the guy wants sell wifi only for &lt;#&gt; and with 3g for &lt;#&gt; . That's why i blanked him."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5824", "copies": 2, "upstream_ids": ["mendeley-5824", "uci-1566"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d:message_type", "suite": "spam", "family": "spamassassin-99be7406a4fa69c66f16", "state": {"channel": "email", "message": "emacs rmail How to sort subject lines and headers.\n\nIn emacs rmail what varieties of different techniques are there for\nsorting your favorite correspondents from the mix?... leaving the hundreds\nof spam commercials. A filter for the campus computer system puts subject\nlines and headers on many of the hundreds of spam commercials\nwith http://spamassassin.org\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01534.0af697b73ae39808dd4035294018be4d", "zefang-4432", "ealvaradob-4239"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2594:message_type", "suite": "spam", "family": "sms_phishing-428ca05849199c671fb1", "state": {"channel": "sms", "message": "No. Did you multimedia message them or e-mail?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2594", "copies": 2, "upstream_ids": ["mendeley-2594", "uci-3875"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3951:is_spam", "suite": "spam", "family": "sms_phishing-5fd86e1563095dab07e3", "state": {"channel": "sms", "message": "I meant as an apology from me for texting you to get me drugs at &lt;#&gt; at night"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3951", "copies": 2, "upstream_ids": ["mendeley-3951", "uci-3621"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing2.mbox#267:is_spam", "suite": "spam", "family": "nazario-ed9f1deeaf161fc2b545", "state": {"channel": "email", "message": "Personal Account Verification\n\nWells Fargo update your account\n\nNew Page 1\n\nDear customers:\n\nWells Fargo is constantly working to increase security for all Online Banking\nusers. To ensure the integrity of our online payment system, we periodically\nreview accounts.\n\nYour account might be place on restricted status. Restricted accounts continue\nto receive payments, but they are limited in their ability to send or withdraw\nfunds.\n\nTo lift up this restriction, you need to login into your account (with your\nusername or SSN and your password), then you have to complete our verification\nprocess. You must confirm your credit card details and your billing information\nas well. All restricted accounts have their billing information unconfirmed,\nmeaning that you may no longer send money from your account until you have\nupdated your billing information on file.\n\nTo initiate the billing update confirmation process, please follow the link\nbellow and fill in the necessary fields:\n\nhttps://online.wellsfargo.com/signon?LOB=CONS\n\nThank you,\n\nWells Fargo - Online Banking\n\nAbout Wells Fargo |\n\nEmployment |\n\nReport Email Fraud |\n\nPrivacy, Security & Legal |\n\nHome\n\n© 2014 - 2024 Wells Fargo. All rights reserved."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#267", "copies": 5, "upstream_ids": ["phishing2.mbox#267", "phishing2.mbox#385", "phishing2.mbox#387", "phishing2.mbox#390", "phishing2.mbox#806"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d:is_spam", "suite": "spam", "family": "spamassassin-3638619e0c52ccda8f04", "state": {"channel": "email", "message": "spamd can't find...\n\nI thought I'd installed razor correctly, but I am seeing the following in my\nlogs. Can anyone give me a hitn?\n\nSep 8 06:46:45 omega spamd[6514]: razor2 check skipped: No such file or\ndirectory Can't locate object method \"new\" via package \"Net::DNS::Resolver\"\n(perhaps you forgot to load \"Net::DNS::Resolver\"?) at (eval 31) line 1,\n<STDIN> line 114. ^I...propagated at\n/usr/lib/perl5/site_perl/5.6.1/Mail/SpamAssassin/Dns.pm line 392, <STDIN>\nline 114.\n\n--\nRobin Lynn Frank\nParadigm-Omega, LLC\nThe only certainty about documentation is that\nwhoever wrote it \"might\" have understood it.\nThe rest of us may not be so lucky.\n\ncell phone? Get a new here for FREE!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01451.b5a50ca35f50e38d37a2eba47399f57d"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3579:message_type", "suite": "spam", "family": "sms_phishing-79f163bb14940aa0e001", "state": {"channel": "sms", "message": "Bears Pic Nick, and Tom, Pete and ... Dick. In fact, all types try gay chat with photo upload call 08718730666 (10p/min). 2 stop texts call 08712460324"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": [0.0, 0.5, 0.5], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3579", "copies": 3, "upstream_ids": ["mendeley-3579", "mendeley-4446", "uci-1374"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2976:message_type", "suite": "spam", "family": "sms_phishing-1b59d2eff7222cfd525f", "state": {"channel": "sms", "message": "Shb b ok lor... Thanx..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2976", "copies": 2, "upstream_ids": ["mendeley-2976", "uci-952"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3:is_spam", "suite": "spam", "family": "spamassassin-8ec0256dc3f10e50aab8", "state": {"channel": "email", "message": "Suggestion - 2 levels of Spam Status\n\nTony L. Svanstrom wrote:\n\nSure - we developers know that - but what I'm talking about is making it\neasier for END USERS to figure out.\n\nWelcome to geek heaven.\nhttp://thinkgeek.com/sf"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01482.a3576ddec34b0481a9a671c1f3c141c3", "zefang-4918", "ealvaradob-4708"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-20:message_type", "suite": "spam", "family": "sms_phishing-1e70c50acc45b170c4d2", "state": {"channel": "sms", "message": "(That said can you text him one more time?)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-20", "copies": 2, "upstream_ids": ["mendeley-20", "uci-1151"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2023#210:is_spam", "suite": "spam", "family": "nazario-140085df0925f1820041", "state": {"channel": "email", "message": "Notice: Password Expiry Notice\n\nFrom Webmail deltafoods.net Server Admin\n\nDear debra\n\nTo help fight spam and improve security services, Please validate your mailbox\n\nValidate northworks.co.uk\n\nSecure Mailbox\n\nThis email was sent to debra to provide an update about your mailbox\n© 2021 | 7/25/2021 7:13:29 a.m."}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#210", "copies": 1, "upstream_ids": ["phishing-2023#210"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429:is_spam", "suite": "spam", "family": "spamassassin-7737fd58f7a70139b991", "state": {"channel": "email", "message": "Sorting\n\nOn Tue, 10 Sep 2021 01:20:56 PDT,\nJ C Lawrence <donna_williams@live.com> wrote:\n\nWhy? What threat does Outlook pose to your organization?\n\nMore detail please? I do use sequences, so I'm familiar with their\nuse, but how can I make indices with them, and how can I keep them\nup to date?\n\nYou're probably running exmh on a local machine. I'm running it on\na very remote machine. In this scenario, the mime handling is weak."}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00990.ee34876c3873d8e6197432ad9c558429"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1357:message_type", "suite": "spam", "family": "sms_phishing-be4197302a95e22df1de", "state": {"channel": "sms", "message": "Of course ! Don't tease me ... You know I simply must see ! *grins* ... Do keep me posted my prey ... *loving smile* *devouring kiss*"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1357", "copies": 2, "upstream_ids": ["mendeley-1357", "uci-5245"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00113-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-168df1f0eed2a93f8a52", "state": {"channel": "email", "message": "Updated payment details - Invoice 4021\n\nInvoice Number: 4021\nAmount Due: €2,450.00\nDue Date: 15 March 2020\nSupplier: Quartz Networks Ltd\n\nElizabeth, please note our bank details have recently changed. Kindly remit the above payment to our new account. Full routing instructions are available here: https://www.quartznetworks-secure.biz/payment-update/inv-4021\n\nThank you for your prompt attention.\nBrian Myers\nAccounts Department\nQuartz Networks Ltd"}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00113-4", "copies": 1, "upstream_ids": ["genu-00113-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from a supplier, says their bank details changed and asks for the next payment to the new account", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74:message_type", "suite": "spam", "family": "spamassassin-494a13a13f153c4350d9", "state": {"channel": "email", "message": "New Medications.\n\nForwarded-by: Rob Windsor <ronald.james@graniteconsulting.io>\nForwarded-by: \"Rich Holland\" <okim@quartzworks.com>\nForwarded-by: Diane <gary86@blueworks.biz>\n\nS t . M o m ' s W o r t\nPlant extract that treats mom's depression by rendering\npreschoolers unconscious for up to six hours.\n\nE m p t y N e s t r o g e n\nHighly effective suppository that eliminates melancholy by\nenhancing the memory of how awful they were as teenagers and\nhow you couldn't wait till they moved out.\n\nP e p t o b i m b o\nLiquid silicone for single women. Two full cups swallowed\nbefore an evening out increases breast size, decreases\nintelligence, and improves flirting.\n\nD u m e r o l\nWhen taken with Peptobimbo, can cause dangerously low I.Q.\ncausing enjoyment of country western music.\n\nF l i p i t o r\nIncreases life expectancy of commuters by controlling road\nrage and the urge to flip off other drivers.\n\nA n t i b o y o t i c s\nWhen administered to teenage girls, is highly effective in\nimproving grades, freeing up phone lines, and reducing money\nspent on make-up.\n\nM e n i c i l l i n\nPotent antiboyotic for older women. Increases resistance to\nsuch lines as, \"You make me want to be a better person ... can\nwe get naked now?\"\n\nB u y a g r a\nInjectable stimulant taken prior to shopping. Increases\npotency and duration of spending spree.\n\nE x t r a S t r e n g t h B u y-O n e-A l l\nWhen combined with Buyagra, can cause an indiscriminate\nbuying frenzy so severe the victim may even come home with a\nDonnie Osmond CD or a book by Dr. Laura.\n\nJ a c k A s s p i r i n\nRelieves the headache caused by a man who can't remember your\nbirthday, anniversary or phone number.\n\nA n t i - t a l k s i d e n t\nA spray carried in a purse or wallet to be used on anyone too\neager to share their life stories with total strangers.\n\nR a g a m e t\nWhen administered to a husband, provides the same irritation\nas ragging on him all weekend, saving the wife the time and\ntrouble of doing it herself."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/01373.55c2a7811975ab50ef0a350992d3ad74"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4:message_type", "suite": "spam", "family": "spamassassin-6fba1e5dc1fec932c83d", "state": {"channel": "email", "message": "Yahoo Finance RSS Beta\n\nGot a stock ticker for which you'd like to have an RSS news feed? Help test the\nbeta RSS feeds we've put up o Yahoo Finance. Take your favorite ticker, say\nYHOO, and put this URL in your news aggregator:..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/01961.b47390d07cd4c2dba35176e9a948bac4"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2223:message_type", "suite": "spam", "family": "sms_phishing-dce3a06cf06f9654716e", "state": {"channel": "sms", "message": "I can send you a pic if you like :)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2223", "copies": 2, "upstream_ids": ["mendeley-2223", "uci-2045"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2553:is_spam", "suite": "spam", "family": "sms_phishing-65ef3cfc44ab3dd8c261", "state": {"channel": "sms", "message": "Call me when u're done..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2553", "copies": 2, "upstream_ids": ["mendeley-2553", "uci-4292"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-966:message_type", "suite": "spam", "family": "sms_phishing-5193b8f2e13e008738d9", "state": {"channel": "sms", "message": "I know you are. Can you pls open the back?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-966", "copies": 2, "upstream_ids": ["mendeley-966", "uci-144"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022:message_type", "suite": "spam", "family": "spamassassin-6d2efd92fc44ad744a12", "state": {"channel": "email", "message": "what is the sound of one knee jerking?\n\nhttp://www.livejournal.com/talkread.bml?journal=jwz&itemid=63598"}, "question": {"type": "choice", "instructions": "Pick the kind of message: legitimate, spam (unwanted bulk or advertising) or phishing (a scam to extract personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-9", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2600:message_type", "suite": "spam", "family": "sms_phishing-2aedeeecef28b8fa71c8", "state": {"channel": "sms", "message": "Am in gobi arts college"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2600", "copies": 2, "upstream_ids": ["mendeley-2600", "uci-2639"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00201-3:is_spam", "suite": "spam", "family": "generated_legit_email-f722df91b05dadce5ecb", "state": {"channel": "email", "message": "Year 4 Trip to the Science Museum\n\nDear Charles,\n\nOur Year 4 trip to the Science Museum is scheduled for Thursday 19 April 2011. Please return the signed consent forms by Tuesday 17 April so we can confirm numbers. More details are available at https://www.crestlabs.net/year4-trip-april2011.\n\nMany thanks,\nBenjamin Baker\nCrest Labs Academy"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00201-3", "copies": 1, "upstream_ids": ["gen-00201-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "school message to parents (trip, closure, parents' evening)"}}
{"id": "spam:zefang_phishing_email:none:zefang-6507:is_spam", "suite": "spam", "family": "zefang_phishing_email-b7da732977f2138c3019", "state": {"channel": "email", "message": "trv notification : ( waha - 06 / 21 / 2012 ) the report named : waha , published as of 06 / 21 / 2012 is now available for viewing on the website ."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-6507", "copies": 2, "upstream_ids": ["zefang-6507", "ealvaradob-6237"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00672-1:is_spam", "suite": "spam", "family": "generated_legit_email-3e8547f4a6f218771fe4", "state": {"channel": "email", "message": "Your November statement is ready\n\nHi Tomasz,\n\nThanks for reaching out earlier. I have checked your account and can confirm that your monthly QuartzLabs credit card statement for October 2022 is now available to view online. The total balance due is £412.55, with a minimum payment of £25.00 required by 20 November 2022.\n\nYou can log in and download your statement securely here: https://www.quartzlabs.de/statements/nov2022\n\nLet me know if you need any further assistance with this.\n\nBest regards,\nJulie Walsh\nQuartzLabs Customer Services"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00672-1", "copies": 1, "upstream_ids": ["gen-00672-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "monthly bank or credit card statement is ready"}}
{"id": "spam:sms_phishing:none:mendeley-3688:is_spam", "suite": "spam", "family": "sms_phishing-f1c724bde3d05d42258f", "state": {"channel": "sms", "message": "Where can download clear movies. Dvd copies."}, "question": {"type": "noul", "instructions": "Is the message spam or phishing, as opposed to a legitimate message? Answer yes or no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-12", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3688", "copies": 2, "upstream_ids": ["mendeley-3688", "uci-1077"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5598:is_spam", "suite": "spam", "family": "sms_phishing-5132f0b18fc49f504745", "state": {"channel": "sms", "message": "Unfortunately i've just found out that we have to pick my sister up from the airport that evening so don't think i'll be going out at all. We should try to go out one of th"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5598", "copies": 2, "upstream_ids": ["mendeley-5598", "uci-1533"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3240:message_type", "suite": "spam", "family": "sms_phishing-bcd2390d069fc4825687", "state": {"channel": "sms", "message": "Hai dear friends... This is my new &amp; present number..:) By Rajitha Raj (Ranju)"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3240", "copies": 2, "upstream_ids": ["mendeley-3240", "uci-5070"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3646:is_spam", "suite": "spam", "family": "sms_phishing-61b1ed89b8e87622feaa", "state": {"channel": "sms", "message": "Thats a bit weird, even ?- where is the do supposed to be happening? But good idea, sure they will be in pub!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3646", "copies": 2, "upstream_ids": ["mendeley-3646", "uci-704"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-8576:is_spam", "suite": "spam", "family": "zefang_phishing_email-67222d108b9ba1ef3e78", "state": {"channel": "email", "message": "global asset operation reorganization in an effort to streamline the corporate center and in recognition of the growing capabilities of enron \u0001 , s regional wholesale business , global asset operations will be reorganized effective 1 may 2013 . our current global asset operations functions will be decentralized and become a business unit responsibility . enron will continue to operate all of its facilities under standardized guidelines and procedures and compliance will be monitored through a formal audit program as in the past . however , our wholesale business units will assume responsibility for operation and management of their specific power plants and other assets . operational energy corporation ( oec ) will become the enron corp \u0001 and john cates will lead the development support activity . john cates \u0001 , organization will continue to provide o & m estimates and due diligence assistance to business units . john will specifically approve o & m estimates for corporate risk assessment decisions . jim noles and this new group will report to the office of the chairman in enron north america . the quality management group ( audit function ) will be assigned to the gas pipeline group ( gpg ) . don hawkins will continue to head quality management and will report to phil lowry , managing director , operations , gpg . in addition to managing the global audit program , don will continue his responsibility for standardized policies and procedures , the crisis management program , iso 9000 certification , the positive impact program , operational reporting , and the asset operations website . the environmental health and safety ( ehs ) group led by henry van will report to mike terraso , enron \u0001 , s chief environmental officer reporting to steve kean . as in the past , henry \u0001 , s group will continue to provide environmental and safety support to the regional business units . enron \u0001 , s wholesale business unit \u0001 , s can now choose from a variety of services provided by oec . these include full turnkey o & m services , o & m estimates , due diligence , and technical support for planned and forced outages . business units are encouraged to take advantage of these services as they operate and maintain the power plants in their businesses . kurt huneke will help transition the asset operations activity as discussed herein , while dedicating his full attention to enron wind company as its chairman and ceo . please join us in wishing these groups success ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-8576", "copies": 2, "upstream_ids": ["zefang-8576", "ealvaradob-8194"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing3.mbox#1082:message_type", "suite": "spam", "family": "nazario-02f9a1bdffd4bcf3877c", "state": {"channel": "email", "message": "Your PayPal account might be compromised\n\nYour PayPal account might be compromised\n\nPayPal\n\nVerisign\n\nProtect Your Password\n\nYou should never give your PayPal password to anyone, including PayPal employees.\n\nYour PayPal account might\nbe compromised\n\nWe have been notified that a card associated with your account has been reported as lost or stolen, or that there were additional problems with your card. Therefore, as a prevention measure, we have temporarely limited access to sensitive PayPal account features.\n\nTo ensure that your account is not compromised please take a review on your recent account history for any unauthorized withdrawals or deposits, and check your account profile to make sure no changes have been made. If any unauthorized activity has taken place on your account, report this to Paypal Security Center immediately.\n\nTo get started, please click the link below:\n\nhttps://www.paypal.com\n\nThis e-mail contains information directly related to your account with us, other services to which you have subscribed, and/or any application you may have submitted. PayPal and its service providers are committed to protecting your privacy and ask you not to send sensitive account information through e-mail.\n\nYou can view our privacy policy and contact information at\n\nhttps://www.paypal.com/cgi-bin/webscr?cmd=p/gen/ua/policy_privacy\n\nPlease understand that this is a security measure intended to help protect you and your account. We apologize for any inconvenience.If you choose to ignore our request, you leave us no choice but to temporary suspend your account. Please do not reply to this e-mail. For assistance, log in to your PayPal account and choose the \"Help\" link in the header of any page.\n\n1410-3573-2309-0292-4887\n\nThank you for using PayPal!\n\nThe PayPal Team\n\nPlease do not reply to this email. This mailbox is not monitored and you will not receive a response. For assistance, log in to your PayPal account and choose the Help link located in the top right corner of any PayPal page.\n\nTo receive email notifications in plain text instead of HTML, update your preferences here.\n\nPayPal Email ID PP059"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing3.mbox#1082", "copies": 1, "upstream_ids": ["phishing3.mbox#1082"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-2766:is_spam", "suite": "spam", "family": "zefang_phishing_email-a8761a8ea8327838ea29", "state": {"channel": "email", "message": "enron / stanford program nick , dinner on sunday would work for me . i shall stay in the bay area till monday morning . vince nick bambos on 09 / 28 / 2022 08 : 33 : 38 pm to : vince . j . kaminski @ enron . com cc : subject : re : enron / stanford program hi vince , i am on the technical program committee of the infocom 2023 conference , and we are meeting in new york city on saturday , october 14 th , to select papers for the conference program . i ' m leaving stanford on friday and getting back on sunday . it might be a possibility to have dinner together on sunday , if that would work for you . in that case i would have to reschedule my flight to land in sfo earlier than i ' m currently scheduled to land . would dinner on sunday work for you ? any chance we can meet monday for lunch ? i look forward to seeing you . best regards , nick vince . j . kaminski @ enron . com wrote : > > nick , > > i shall be in stanford oct 14 - 15 , visiting my family . > i would be glad to meet you ( and possibly giuseppe - your call ) for lunch . > please , let mer know if you are free on one of these days . saturday would > work better for me . > > vince > > nick bambos on 09 / 21 / 2022 02 : 09 : 46 pm > > to : stinson . gibner @ enron . com > cc : vince . j . kaminski @ enron . com > subject : re : enron / stanford program > > stinson , > > great ! i ' m looking forward to a very productive collaboration . > i ' ll immediately start doing giuseppe ' s papers , for him to work > on the enron / stanford program . > > many thanks to you and vince , and i hope to see you soon at stanford > or enron . if i remember correctly , vince is visiting stanford in > october . > > best regards , > > nick > > stinson . gibner @ enron . com wrote : > > > > nick , > > > > i spoke with paul racicot , head of trading for ebs , north america this > > morning . he said that he is happy to send the $ 100 , 000 for your program > > from his budget . i have forwarded to him the draft letter to accompany > > the funds and will try to follow up to make sure that the money is sent > > promptly . > > > > - - stinson"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-2766", "copies": 2, "upstream_ids": ["zefang-2766", "ealvaradob-2655"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-843:is_spam", "suite": "spam", "family": "sms_phishing-fd134e5029b7a7deb613", "state": {"channel": "sms", "message": "I'm really not up to it still tonight babe"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-843", "copies": 2, "upstream_ids": ["mendeley-843", "uci-88"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00221-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-ed706e039c5f39b00c0f", "state": {"channel": "email", "message": "Action Required: StreamVault subscription renewal unsuccessful\n\nDear Freya Rossi,\n\nI hope this email finds you well. I am writing to you from the billing department at StreamVault regarding your premium cloud storage subscription. We attempted to process your annual renewal payment of $119.88 earlier today, 22 February 2023, but unfortunately, the transaction was declined by your card issuer. Because your subscription has now entered a grace period, we want to ensure you do not experience any interruption to your service.\n\nStreamVault provides continuous backup and synchronization for over forty thousand customers worldwide, and we take the security of your stored files very seriously. When a payment fails, our system automatically restricts upload privileges to prevent further data accumulation on an inactive plan. However, all your existing files remain safe and accessible in read-only mode while we work together to resolve this billing issue.\n\nTo avoid any permanent changes to your account status, we kindly ask you to update your payment method and re-authorize the renewal charge. Please note that if the balance remains unsettled by 8 March 2023, your account will be downgraded to our free tier, and any data exceeding the two-gigabyte limit may be scheduled for archival deletion in accordance with our terms of service. We certainly want to help you avoid losing access to your important documents, photographs, and project files.\n\nUpdating your billing information is quick and fully encrypted. Simply click the secure link below to visit our payment verification portal, where you can enter your new card details or choose an alternative payment method such as direct bank transfer. The entire process takes less than two minutes.\n\nUpdate Payment Details Now\nhttps://www.silverconsulting.verify-id.com/update-payment/freya-rossi\n\nOnce your payment has been successfully processed, you will receive an automated confirmation receipt via email, and your premium features will be restored instantly. If you recently changed banks or received a replacement card, updating your information now will also prevent future billing disruptions.\n\nShould you need any assistance or have questions about your invoice, please reply directly to this message, and a member of our customer care team will be happy to guide you through the process. We value your continued trust in StreamVault and are committed to keeping your data secure.\n\nThank you for your prompt attention to this matter.\n\nSincerely,\nAhmed White\nBilling Support Specialist\nStreamVault Cloud Services\nSilver Consulting Inc."}, "question": {"type": "noul", "instructions": "Is this message spam, or an attempt to phish personal details or money? Answer no if it is just a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-15", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00221-3", "copies": 1, "upstream_ids": ["genu-00221-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a subscription renewal failure notice that pretends to be from a streaming, software or cloud storage service and asks for new payment details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00709-3:message_type", "suite": "spam", "family": "generated_legit_email-fe9bcb2bf2cad9378340", "state": {"channel": "email", "message": "Commercial Liability Insurance Renewal - Action Required\n\nStephen,\n\nThanks for flagging this earlier today. I have just reviewed the commercial liability insurance renewal documents sent over by our broker, and we need to act quickly to avoid any lapse in coverage for River Design.\n\nOur current policy expires on 15 July 2024. The renewal premium has increased slightly to $14,250.00 annually, which reflects the additional staff we hired in Q1. I have uploaded the full breakdown and the updated terms to our internal shared drive so the management team can review them before signing off. You can access the policy documents here: https://www.riverdesign.net/internal/insurance-renewal-2024\n\nCould you please review the coverage limits on pages 3 and 4? Specifically, I want to make sure the professional indemnity clause still adequately covers the new municipal contracts we took on in May. If everything looks correct, we need to submit the signed renewal forms and authorize the wire transfer no later than Friday, 8 July 2024.\n\nIf you spot any discrepancies or feel we should negotiate the premium increase, let me know by end of day Thursday so I can get our broker on a call before the weekend.\n\nWe cannot afford any gap in coverage, especially with the site inspections coming up next month.\n\nThanks,\nSusan Thompson\nOperations Manager\nRiver Design LLC"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00709-3", "copies": 1, "upstream_ids": ["gen-00709-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "insurance renewal notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6:message_type", "suite": "spam", "family": "spamassassin-764cbefcf88259ab9a12", "state": {"channel": "email", "message": "Another bug\n\nChris said:\n\nAfter sending the report I started fiddling with the Ftoc_RescanLine\nstuff and I havn't seen it since. I can't really tell if it was my\nhacking that removed it, or if it was some transitional magic\nhappening since this was the first invocation of that version of\nexmh. I've backed out my stuff now, and'll let you know if it happens\nagain.\n\nOn another thing with the Ftoc_RescanLine stuff. This routine is\ncalled at times still unclar to me. An inspection of the routine\nsuggests that it is used in the transition of a message to/form\n\"current\" state to re-paint the ftoc line. However, checking the\nmsg.tcl code (MSgChange) we find:\n\nif {$msgid != {}} {\n# Allow null msgid from Msg_ShowWhat, which supplies line\ninstead\nif {$msgid < 0} return\n} else {\n# line null too, try using first in folder\nif {[string length $line] == 0} { set line 1 }\nset msgid [Ftoc_MsgNumber [Ftoc_FindMsg $msgid $line]]\n}\nFtoc_ClearCurrent\nMh_SetCur $exmh(folder) $msgid\nFtoc_ShowSequences $exmh(folder)\n\nThe Ftoc_ClearCurrent calls Ftoc_RescanLIne to clear the '+' sign\n_before_ the on-disk transition is made. I hacked this stuff and more\nor less changed the order. It works, but fails on some folder changes.\n\nMy feeling of this is that we really do not want to have a\nFtoc_ClearCurrent, but rather just a RescanLine and the caller had\nbetter know the line or msgid. That routine shoud just put in the '+'\nif the line/msg in question happened to be the cur msg. Thoughts?\n\n/Anders"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00687.8b31d13e4d04a9e78a7e01b491f6d2f6"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1449:message_type", "suite": "spam", "family": "sms_phishing-34f401c12f235332dd89", "state": {"channel": "sms", "message": "S...from the training manual it show there is no tech process:)its all about password reset and troubleshooting:)"}, "question": {"type": "choice", "instructions": "What type of message is this: legitimate, spam, or phishing? Spam means unwanted bulk or advertising, phishing means a scam for personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-5", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1449", "copies": 2, "upstream_ids": ["mendeley-1449", "uci-2132"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-12695:is_spam", "suite": "spam", "family": "zefang_phishing_email-304103b6bcb57c993a80", "state": {"channel": "email", "message": "mutually agreed upon changes okay larry , i have your fax and the message . unfortunately the corporate policy is very clear and standard on these matters and we cannot make exceptions . it would appear impossible for us to proceed with the non - disclosure agreement if it is not consistent with the version suggested by our legal department . please let me know your thoughts . thanks . rakesh"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-12695", "copies": 2, "upstream_ids": ["zefang-12695", "ealvaradob-12016"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4674:message_type", "suite": "spam", "family": "sms_phishing-ac571105410457850beb", "state": {"channel": "sms", "message": "How long does applebees take"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4674", "copies": 1, "upstream_ids": ["mendeley-4674"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3919:message_type", "suite": "spam", "family": "sms_phishing-1998f66696a3aa04d861", "state": {"channel": "sms", "message": "But my family not responding for anything. Now am in room not went to home for diwali but no one called me and why not coming. It makes me feel like died."}, "question": {"type": "choice", "instructions": "Decide which kind this message is: legitimate; spam, meaning unwanted bulk or advertising; or phishing, meaning a scam after personal details, passwords or money.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-17", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3919", "copies": 2, "upstream_ids": ["mendeley-3919", "uci-2744"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3958:message_type", "suite": "spam", "family": "sms_phishing-d4792c3bcd2d04cf83af", "state": {"channel": "sms", "message": "That's the trouble with classes that go well - you're due a dodgey one … Expecting mine tomo! See you for recovery, same time, same place"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3958", "copies": 2, "upstream_ids": ["mendeley-3958", "uci-3971"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1656:message_type", "suite": "spam", "family": "sms_phishing-f769e454e88c79810929", "state": {"channel": "sms", "message": "Aaooooright are you at work?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1656", "copies": 2, "upstream_ids": ["mendeley-1656", "uci-156"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-301:message_type", "suite": "spam", "family": "sms_phishing-eee5d16d0b2845d6966a", "state": {"channel": "sms", "message": "FR'NDSHIP is like a needle of a clock. Though V r in d same clock, V r nt able 2 met. Evn if V meet,itz only 4few seconds. Bt V alwys stay conected. Gud 9t;-)"}, "question": {"type": "choice", "instructions": "Say what kind of message this is: legitimate, spam (unwanted bulk or advertising), or phishing (a scam aiming to get personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-19", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-301", "copies": 2, "upstream_ids": ["mendeley-301", "uci-1086"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1923:message_type", "suite": "spam", "family": "sms_phishing-453ec0f39cf83a95a4be", "state": {"channel": "sms", "message": "Hello lover! How goes that new job? Are you there now? Are you happy? Do you think of me? I wake, my slave and send you a teasing kiss from across the sea"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1923", "copies": 2, "upstream_ids": ["mendeley-1923", "uci-1487"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2025#180:is_spam", "suite": "spam", "family": "nazario-ec069b02fcfc1f678cf0", "state": {"channel": "email", "message": "[CASE ID: #BGE-xijlxppvebu: Update Service Terms and Condtions on oaktech.net To Avoid matthew45@btinternet.com being terminated\n\nUPDATED SERVICE TERMS\n\nAttn Olga jeffrey\n\nWe hereby regret to inform you that matthew45@btinternet.com has not approved our new terms/conditions and will be deleted from prairietech.net today 2017/4/28 22:28:19\nincluding all information data on matthew45@btinternet.com\n\nSee below and log-in to get an upgraded mailbox to avoid being discontinued from our services an future offers.\n\nUPDATE TERMS\n\nmeridianfoods.co.uk Services.\n\nThank you for using horizonfoods.net\n\nDisclaimer:\n\nThis email and any attachment there of are confidential and privileged to matthew45@btinternet.com. If you have received it in error, please delete immediately and notify the sender. Do not disclose, copy, circulate or in any way use. The information contained therein is for the intended address(es)."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#180", "copies": 1, "upstream_ids": ["phishing-2025#180"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2551:is_spam", "suite": "spam", "family": "sms_phishing-a3f8e03d0ac9c1ef8d67", "state": {"channel": "sms", "message": "Hi baby ive just got back from work and i was wanting to see u allday! I hope i didnt piss u off on the phone today. If u are up give me a call xxx"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2551", "copies": 2, "upstream_ids": ["mendeley-2551", "uci-4133"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3430:is_spam", "suite": "spam", "family": "sms_phishing-d890ed08f99cfea625d5", "state": {"channel": "sms", "message": "Promotion Number: 8714714 - UR awarded a City Break and a £200 Summer Shopping spree. 08718738034"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3430", "copies": 1, "upstream_ids": ["mendeley-3430"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1453:message_type", "suite": "spam", "family": "sms_phishing-4f191526faf0cd5773eb", "state": {"channel": "sms", "message": "We tried to contact you re our offer of New Video Phone 750 anytime any network mins HALF PRICE Rental camcorder call 08000930705 or reply for delivery Wed"}, "question": {"type": "choice", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "none", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1453", "copies": 3, "upstream_ids": ["mendeley-1453", "mendeley-5599", "uci-839"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00265-0:is_spam", "suite": "spam", "family": "generated_legit_email-7313c99b7367f861f702", "state": {"channel": "email", "message": "Planned IT Maintenance – Beacon Foods Network\n\nBeacon Foods IT Department\nNotice: Scheduled System Maintenance\n\nDear Ryan James,\n\nPlease be advised that the IT department will perform planned maintenance on our internal servers and email systems.\n\nMaintenance window: Saturday 27 December 2015, 22:00 to Sunday 28 December 2015, 06:00 GMT.\n\nDuring this time, access to the shared drives, internal webmail, and the intranet portal will be unavailable. Please save all work and log off before 22:00 on Saturday.\n\nRead the full maintenance schedule\nhttps://www.beaconfoods.com/it/maintenance-dec2015\n\nIf you experience issues after the maintenance window closes, please call the helpdesk at 01632 960 441.\n\nRegards,\nChloe Hughes\nIT Support Manager\n\nBeacon Foods Ltd\nUnit 4, Treforest Industrial Estate\nPontypridd CF37 5YF\nPrivacy policy: https://www.beaconfoods.com/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00265-0", "copies": 1, "upstream_ids": ["gen-00265-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "IT department notice about planned maintenance"}}
{"id": "spam:nazario:none:phishing2.mbox#97:message_type", "suite": "spam", "family": "nazario-30ae9c50c652ed1349d2", "state": {"channel": "email", "message": "Notification of Limited Account Access - Case ID : PP-049-681-317\n\nNotification of Limited\nAccount Access\n\nDear PayPal Valued Costumer,\n\nPayPal is committed to maintaining a safe environment for\nits community of buyers and sellers. To protect the security\nof your account, PayPal employs some of the most advanced\nsecurity systems in the world and our anti-fraud teams\nregularly screen the PayPal system for unusual activity.\n\nWe are contacting you to remind you that on Nov.\n29, 2011 our Account Review Team identified some\nunusual activity in your account. In accordance with PayPal's\nUser Agreement and to ensure that your account has not been\ncompromised, access to your account was limited. Your account\naccess will remain limited until this issue has been\nresolved.\n\nIn order to secure your account and\nquickly restore full access, we may require you to\nverify or update your Personal\nInformation.\nIf you choose to ignore our request, you leave\nus no choice but to temporary suspend your account.\n\nClick\nhere to login and restore your account\naccess\n\nCase ID Number: PP-049-681-317\nThis is a final\nreminder to log\nin to PayPal as soon as possible.\n\nThank you for your prompt attention to this matter. Please\nunderstand that this is a security measure meant to help\nprotect you and your account.\nWe apologize for any\ninconvenience.\n\nSincerely,\nPayPal Account Review Department\nPayPal\nEmail ID PP773\n\nPlease do not reply to this e-mail.\nMail sent to this address cannot be answered. For\nassistance, log\nin to your PayPal account and choose the \"Help\" link\nin the footer of any page.\n\nPayPal Email ID\nPP773\n\nProtect\nYour Account Info\n\nMake sure you never\nprovide your password to fraudulent websites.\n\nTo safely and securely access the PayPal\nwebsite or your account, open a new web browser\n(e.g. Internet Explorer or Netscape) and type in\nthe PayPal URL (https://www.paypal.com/row/) to be\nsure you are on the real PayPal\nsite.\n\nPayPal will never ask you to enter\nyour password in an email.\n\nFor more\ninformation on protecting yourself from fraud,\nplease review our Security Tips at\nhttps://www.paypal.com/row/securitytips\n\nProtect\nYour Password\n\nYou should never give your PayPal\npassword to anyone, including PayPal\nemployees."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#97", "copies": 2, "upstream_ids": ["phishing2.mbox#97", "phishing2.mbox#98"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2538:message_type", "suite": "spam", "family": "sms_phishing-440b6ce07066ae5c2d8a", "state": {"channel": "sms", "message": "No..but heard abt tat.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2538", "copies": 2, "upstream_ids": ["mendeley-2538", "uci-1001"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3719:message_type", "suite": "spam", "family": "sms_phishing-ff6b60d93aef4d2a2b02", "state": {"channel": "sms", "message": "Honeybee Said: *I'm d Sweetest in d World* God Said: U Havnt Met d Person Reading This Msg"}, "question": {"type": "choice", "instructions": "Is this message legitimate, spam (unwanted bulk or advertising), or phishing (a scam after personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3719", "copies": 1, "upstream_ids": ["mendeley-3719"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5090:is_spam", "suite": "spam", "family": "zefang_phishing_email-2af116403fa8be3720c9", "state": {"channel": "email", "message": "a heartwarmer : just imagine ~ welcome to heartwarmers ~ http : / / www . heartwarmers . com the best thing to happen to mornings since the sun ! your morning thought for the day : retirement at 65 is ridiculous . when i was 65 i still had pimples . - - george burns it ' s summer vacation time and parents will soon be hearing from their kids that they ' re bored . one of the biggest regrets we hear from the older generation , is that kids lack the playful imagination that was common a generation or two ago . back in the \" olden days \" , kids knew how to entertain themselves with surprisingly little . there was no such thing as playstation . kids ran outside to play kick - the - can in a game that was shockingly not organized by adults . they actually did it themselves . today , mike shares some thoughts about kids and modern imagination - - or lack thereof . . . what do you think ? support our sponsors they keep our service priceless . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ hundreds of our readers are saving a small fortune . . . heartwarmer , petwarmer and kidwarmer subscribers are discovering the tremendous benefits of using tel 3 advantage to make long distance and international calls - - some saving 90 % ! get started now by calling : 800 - 330 - 6897 . when asked for your \" activation code \" say : 311 - 673 . or you can click here to get complete details and sign up online : http : / / www . pocketwarmers . com / c _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ get free software for your computer hundreds of titles to choose from . you only pay postage . in fact , order 4 disks and get free postage for the 5 th . http : / / www . pocketwarmers . com / d just imagine by michael t . smith when i married ginny , i gained a new son , daughter and four grandsons . i became an instant grandpa . over the july 4 th holiday , i got to meet my new daughter and her three boys for the first time . it took a day for my new grandsons to warm up to me . the first day they stared at me , perhaps wondering what they were supposed to do with this man , who they were told is \" grandpa . \" i waited patiently . i knew boys like to play . they ' d come to me in their own time . on day two , the oldest two were doing summersaults over my lap . the youngest , ben , took a little longer , but yesterday he raced me across the yard thirty times and won every time . the weekend brought back memories of my own childhood . we had such imaginations . in our minds , a tree was a tower to spot approaching bad guys , a big rock became a mountain , a fallen log a space ship headed to the stars . we had toys , but we had to use our imagination , not like the new toys i see my grandkids with . our toys didn ' t talk , and if they did , you pulled a string to make it work . we had blocks to build , crayons to create , trucks and cars to push . they were simple toys that required imagination . i thought about the toys in my garage , the ones i saved , when my children outgrew them . they were simple and needed their imagination to work . mr . potato head allowed them to learn parts of the body and giggle at the funny face with an ear where its mouth should have been . i still have an old plastic phone . it has a dial to turn , a bell that rings when you push a button , and that ' s it . it doesn ' t talk , squawk , beep or move around the room . it ' s simple and was used when they played house , nurse , doctor , and secretary . i have a toy doctor ' s bag , with all the plastic doctor ' s tools . i remember all the broken bones , cuts , bangs , scratches and aches my daughter repaired as i lay in her office moaning . most toys today do it all . we don ' t need imagination - - it comes in a box . video games that take us into another world or reality , talking toys , with vocabularies better than most people , computerized toys to teach the alphabet . kids sit and have imagination brought to them . the teaching toys are great , but children tire of them - - it ' s like being in school . when my son was young and tired of his toys , he would come to me and say , \" dad , i ' m bored . \" \" go find something to do , \" i ' d reply . \" i don ' t know what to do . \" \" go outside and find a friend . \" \" naaa ! i don ' t want to do that . \" \" use your imagination . \" \" huh ? \" he asked . \" well , when i was young . . . \" you know the routine . sports is the same . we ' re entertained for hours watching someone else have fun doing what they ' re good at . wouldn ' t it be better to be playing yourself ? if we don ' t like the sports available to watch , we create new ones . it doesn ' t matter what , people will pay to see it , because they need to be entertained . i haven ' t heard of world championship worm digging , but if someone offered a large cash price , people would buy tickets to scream at the contestants . i worry . are we becoming a society that needs outside influence to have fun ? i think i ' ll go climb a tree . maybe there ' s a pirate ship on the horizon or someone is attacking my castle . just imagine . - - michael t . smith michael lives in new jersey with his new wife and son . from our mailbag : ( in response to last week ' s story , cinderella men ) dear heartwarmers : last week ' s heartwarming story opened a door to memory lane . i too , was born in the 30 s and felt some of the poverty of that time . i remember my mom and dad who never complained of the fatigue that was so ever present in their lives . we were very fortunate in as much as we always had plenty of food and warm shelter . we had a garden and salvaged every edible product available - - even certain \" weeds \" that were quite tasty after mother got all the spices in them . i had several beautiful dresses made from feed sacks that i wore with pride ! we were very poor , but so rich indeed because we had love and the knowledge that god was upper most in our lives , and it taught us lessons that had never been printed in the books . i deeply appreciate those"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5090", "copies": 3, "upstream_ids": ["zefang-5090", "zefang-6904", "ealvaradob-4874"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": {"original_chars": 11005, "kept_chars": 5996}, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:nazario:none:phishing-2025#88:message_type", "suite": "spam", "family": "nazario-6463f766127d3b9cd6eb", "state": {"channel": "email", "message": "Webmail account validation for vistapartners.biz user(s)\n\nVerify your email address\nHello Carlos,\n\nTo continue using your webmail account (rrivera@comcast.net ( ) ),\nplease verify that this is your email address.\n\nVerify email address\n( https://xwebmail-bautho-xi.vercel.app/2096/cpsess4535115678/roundcube/user-validation#am9zZUBtb25rZXkub3Jn )\n\nThis link will expire in 3 days. You might lose access to your\nmailbox if action is not taken immediately.\nFor help, contact us through our Help center\n( https://roundcube.net ) ."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2025#88", "copies": 1, "upstream_ids": ["phishing-2025#88"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1780:message_type", "suite": "spam", "family": "sms_phishing-a5ac5c5038c2d8777335", "state": {"channel": "sms", "message": "Hey ! Don't forget ... You are MINE ... For ME ... My possession ... MY property ... MMM ... *childish smile* ..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1780", "copies": 2, "upstream_ids": ["mendeley-1780", "uci-5289"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-648:is_spam", "suite": "spam", "family": "zefang_phishing_email-6282f3b1a7be0355c7dc", "state": {"channel": "email", "message": "John P. Looney wrote:\nRansom is a pint.Pádraig.\n--"}, "question": {"type": "noul", "instructions": "Is this message unwanted spam or a phishing scam? Yes for either; no if it is an ordinary message someone would want.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-10", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-648", "copies": 2, "upstream_ids": ["zefang-648", "ealvaradob-626"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-144:is_spam", "suite": "spam", "family": "sms_phishing-c2b17a44a5337a8ca836", "state": {"channel": "sms", "message": "Jan is fast approaching. So, Wish u a very Happy New Year Happy Sankranti Happy republic day Happy Valentines Day Happy Shivratri Happy Ugadi Happy Fools day Happy May Day Happy Independence Day, Happy Friendship,Mother,Father,Teachers,Childrens Day, &amp; HAPPY BIRTHDAY 4 U. Happy Ganesh festival Happy Dasara Happy Diwali Happy Christmas &lt;#&gt; Good Mornings Afternoons, Evenings Nights. RememberI AM the first to WISHING U ALL THESE...your's Raj"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-144", "copies": 1, "upstream_ids": ["mendeley-144"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-562:is_spam", "suite": "spam", "family": "sms_phishing-b3c5f4606a5e31bc863d", "state": {"channel": "sms", "message": "Come to mu, we're sorting out our narcotics situation"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-562", "copies": 2, "upstream_ids": ["mendeley-562", "uci-271"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2016#426:is_spam", "suite": "spam", "family": "nazario-00735d2e1e7b57b8212e", "state": {"channel": "email", "message": "Warning*** Upgrade your mailbox\n\nServer Message\n\nDear lisa19@verizon.net Your Mailbox capacity is almost full, you have used 14.5 GB/ 15GB\n\n; Kindly upgrade your Mailbox to the latest version to get unlimited access\n\nupgrade\n\nAnd avoid suspension of your mailbox, upgrade is free. failure to upgrade your account will result to termination of your account. Regards.\nEmail Administrator\n\nThis message is auto-generated from E-mail security server, and replies sent to this email can not be delivered.\nThis email is meant for:=lisa19@verizon.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#426", "copies": 1, "upstream_ids": ["phishing-2016#426"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00737-2:is_spam", "suite": "spam", "family": "generated_legit_email-f872a15a3c2ce2b38340", "state": {"channel": "email", "message": "Order Confirmation - Autumn Craft Fair Supplies\n\nOrder number: QD-2009-8834\nAmount: £42.50\nPayment method: Visa ending 9012\nItems: 5x Canvas Tote Bags, 10x Acrylic Paint Sets\n\nView your receipt here: https://www.quartzdesign.biz/orders/QD-2009-8834\n\nThank you for supporting the Riverside Community Art Group!\nEmma Kowalski"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00737-2", "copies": 1, "upstream_ids": ["gen-00737-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:zefang_phishing_email:none:zefang-3907:is_spam", "suite": "spam", "family": "zefang_phishing_email-133aac88c85b7c3748b0", "state": {"channel": "email", "message": "textile charset = gb 2312 \" > dear sir / madam , we are a manufacturer of jackets , fleece , shirts , polo and t - shirts . we would like to establish business cooperation with your company . hereby we introduce some of our running stock styles for your consideration . garments production factory : texwin garments factory industrial development zone , hengsan , guangning city , guangdong province , china tel : ( 86 ) 758 8651529 fax : ( 86 ) 758 8651342 e . mail : info @ max - po . comwebsite : http : / / www . tex - po . com / we have the following commitments for our running stocks : 1 . our products are qualified to the iso standards and aql standard . 2 . our price level is most competitive and realistic . 3 . our ordered quantity could be flexible from 50 pcs per colour per style . 4 . our products can be with your own labels , printings and embroideries . 5 . we can provide immediate delivery upon your order confirmation . website : www . tex - win . com / www . tex - po . com / www . texcotextile . com / www . sailing - po . com / www . hunting - po . com . quality assurance : 1 . workmanship and garments finishing standard pass the aql 2 . 5 standard approved by sgs 2 . dye - stuff passed the azo - free standard 3 . ph value passed the ( iso 307 - 2008 ) standard 4 . dimensional stability to washing passed the ( iso 6330 - 2012 ) standard 5 . color fastness to washing passed the ( iso / 105 co 6 a 2 s - 2022 ) standard 6 . color fastness to rubbing passed the ( iso / 105 xl 2 - 2021 ) standard 7 . pre - shrinkage finished 8 . 9 . ref . no . gto 03 ( leisure shirt ) ref . no . gto 02 ( leisure shirt ) ref . no . gto 01 ( adventure shirt ) ref . no . fgo 01 ( promotional polarfleece gilet ) casual cargo trousers shorts ref . no . cto 01 ( functional coolmax polo ) besides , we could also produce above garment items and we could also produce styles with your designs . we would make customized logo design in embroideries , imprints , labels and badges on the garments according to your design . for more information , please visit our hereafter websites or contact us by info @ max - po . com . contact : contact person : mr . christopher ho ( hk mob : ( 852 ) 9345 5790 cn mob : ( 86 ) 132 46472480 ) mr . terry choi hunting wears tracksuits t - shirts waterproof wears fishing wears motor racing wears baby wears fleece wears ref . no . njo 01 ( promotional windbreaker ) anti - pilling finished for polarfleece zipper can be qualified by oeko ¨ c standard in durability and dye - stuff"}, "question": {"type": "noul", "instructions": "Check the message: is it spam or phishing? Say yes for unwanted advertising, bulk sends or scams trying to get personal details or money; say no for a genuine message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-7", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3907", "copies": 2, "upstream_ids": ["zefang-3907", "ealvaradob-3744"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1113:message_type", "suite": "spam", "family": "sms_phishing-b375edd3404e4831c827", "state": {"channel": "sms", "message": "You won't believe it but it's true. It's Incredible Txts! Reply G now to learn truly amazing things that will blow your mind. From O2FWD only 18p/txt"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1113", "copies": 3, "upstream_ids": ["mendeley-1113", "uci-3574", "uci-5037"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00185-2:is_spam", "suite": "spam", "family": "generated_legit_email-365e9e898486158dac6f", "state": {"channel": "email", "message": "Summer Term Registration Now Open\n\nNORTH PARTNERS COMMUNITY LEARNING\n\nSummer Term Registration Now Open\n\nDear Diane King,\n\nWe are excited to announce that registration for our Summer 2008 community courses is officially open. As a valued member of our learning network, you get priority enrollment.\n\nThis term we are offering:\n- Introduction to Digital Photography\n- Community First Aid Certification\n- Conversational Spanish (Beginner)\n- Local History Walking Tours\n\nClasses begin on 1 August 2008 and run through mid-September. Spots are limited, so please secure your place soon.\n\nRegister Now\nhttps://www.northpartners.com/courses/summer-2008\n\nIf you have questions about scheduling or fees, contact us at info@northpartners.com.\n\nThank you for supporting North Partners Community Learning.\n\nWarm regards,\nRahul Ross\nProgram Director\n\n---\nNorth Partners Community Learning\n42 Harbour Road, Bristol BS1 4RN, United Kingdom\nPrivacy Policy: https://www.northpartners.com/privacy"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00185-2", "copies": 1, "upstream_ids": ["gen-00185-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "online course or university notice"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8:is_spam", "suite": "spam", "family": "spamassassin-5297eebe927a70921a50", "state": {"channel": "email", "message": "fully-public corpus of mail available\n\n(trimmed cc list)\n\nDaniel Quinlan said:\n\nI'll note that in the README.\n\nOK; maybe rewriting the message-ids will help here, that should allow\nus to pick them out. I'll do that.\n\nnah, me neither.\n\nYep. The primary reason I released this, was to provide a good, big\ncorpus for academic testing of filter systems; it allows results to\nbe compared between filters using a known corpus.\n\nFor SpamAssassin development, everyone has to maintain their own corpus.\n\n--j."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/01546.600ca62f96dca0db15aa9ebbb19426a8", "zefang-5793", "ealvaradob-5550"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a:is_spam", "suite": "spam", "family": "spamassassin-71c6ec1b384066b84274", "state": {"channel": "email", "message": "CVS report\n\nJust cvs up'ed and nowadays Catch-up Unseen is __extremely__ slow on\nlarge (>100 msgs) unseen sequences. Anybody else having this problem?\n\n/A"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a", "copies": 3, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00788.4fe7ac37abaf058662d024d6c482119a", "zefang-4267", "ealvaradob-4085"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4140:message_type", "suite": "spam", "family": "sms_phishing-eb0b4b60cc9fc36e6bb6", "state": {"channel": "sms", "message": "Boy you best get yo ass out here quick"}, "question": {"type": "choice", "instructions": "Classify the message as legitimate, spam, or phishing. Spam is unwanted bulk or advertising; phishing is a scam trying to get personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-1", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4140", "copies": 2, "upstream_ids": ["mendeley-4140", "uci-3069"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5746:message_type", "suite": "spam", "family": "sms_phishing-a893edee8e1e34d2bd0e", "state": {"channel": "sms", "message": "Now press conference da:)"}, "question": {"type": "choice", "instructions": "Sort this message into legitimate, spam or phishing. Phishing tries to get personal details, passwords or money; spam is unwanted bulk or advertising.", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-4", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5746", "copies": 2, "upstream_ids": ["mendeley-5746", "uci-1543"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00019-0:is_spam", "suite": "spam", "family": "generated_legit_email-e2a2fc875e2ab7416d5d", "state": {"channel": "email", "message": "Your Cedar Foods Order Receipt #CF-2019-8834\n\nHi Shirley,\n\nThank you for your recent purchase from Cedar Foods. We have received your order and payment successfully. Below is a summary of your transaction for your records.\n\nOrder Details:\n- Order Number: CF-2019-8834\n- Date of Purchase: 20 February 2019\n- Customer Email: mateo31@btinternet.com\n\nItems Ordered:\n1. Organic Quinoa Blend (500g) - £6.50\n2. Cedar Roasted Almonds (250g) - £4.99\n3. Cold-Pressed Olive Oil (1L) - £12.75\n4. Dried Apricots (300g) - £3.80\n\nSubtotal: £28.04\nShipping: £4.50\nTotal Charged: £32.54\n\nNext Steps:\n1. Our warehouse team is currently packing your items.\n2. You will receive a separate shipping confirmation email with tracking details within 24 hours.\n3. Please ensure someone is available to receive the delivery by Tuesday, 25 February 2019, as the olive oil requires a signature upon arrival.\n\nIf you need to make any changes or cancel this order, you must do so before 5:00 PM GMT on Friday, 21 February 2019. After that deadline, the order will be dispatched and cannot be altered.\n\nYou can view your full order history and download your invoice here: https://www.cedarfoods.net/orders/CF-2019-8834\n\nIf you have any questions about your receipt or delivery, please reach out to our customer support team at +44 20 7946 0123. We are available Monday through Saturday, 8 AM to 6 PM.\n\nThank you for choosing Cedar Foods!\n\nBest regards,\nTyler Robinson\nCustomer Service Team\nCedar Foods\nsupport@cedarfoods.net"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00019-0", "copies": 1, "upstream_ids": ["gen-00019-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "order receipt from an online shop"}}
{"id": "spam:generated_legit_email:none:gen-00079-5:is_spam", "suite": "spam", "family": "generated_legit_email-219d293bbc71a99df6ef", "state": {"channel": "email", "message": "Community Pantry Delivery Missed - Redelivery Options\n\nHi Ahmed,\n\nI stopped by your house at 142 Elm Street this morning around 10:30 AM to drop off your weekly share from the Pioneer Health community food pantry, but it looks like nobody was home. I did not want to leave the fresh produce and frozen items on the porch since the weather is getting quite warm today and we do not want anything spoiling.\n\nI have taken your box back to our distribution center at the Pioneer Health building on 5th Avenue. We can try delivering it again tomorrow, Sunday, October 15, anytime between 9:00 AM and 1:00 PM. Alternatively, you are welcome to swing by and pick it up yourself before 5:00 PM today.\n\nPlease give me a call or send a quick reply to this email to let me know what works best for you so I can update our driver route sheet. We want to make sure you get your groceries while everything is still fresh.\n\nTake care,\nSandra Murphy\nVolunteer Delivery Coordinator\nPioneer Health Community Outreach"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00079-5", "copies": 1, "upstream_ids": ["gen-00079-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:sms_phishing:none:mendeley-2206:is_spam", "suite": "spam", "family": "sms_phishing-15f94f814f29479e4282", "state": {"channel": "sms", "message": "Raji..pls do me a favour. Pls convey my Birthday wishes to Nimya."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2206", "copies": 1, "upstream_ids": ["mendeley-2206"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_unwanted_email:none:genu-00089-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-d6c48a2dca7404bdf2f8", "state": {"channel": "email", "message": "Emergency Relief Fund - Card Verification Needed\n\nHi Ananya,\n\nI am following up on my previous message regarding the emergency earthquake relief effort in Turkey and Syria. Union Works is actively collecting donations to support the families who have lost everything, and we need your help to finalize the transaction.\n\nTo proceed with your generous pledge, please provide your debit or credit card details by replying to this email, or use our secure submission form here: https://www.myunionworks-online.com/relief-donate-verify\n\nWe must receive your information by 19 April 2018 to include your contribution in this week's wire transfer to the rescue teams on the ground. Your support means the world right now.\n\nWarmly,\nYuki Petrov\nCommunity Outreach Director\nUnion Works"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00089-3", "copies": 1, "upstream_ids": ["genu-00089-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-2344:message_type", "suite": "spam", "family": "sms_phishing-e2c955c99279442d66b4", "state": {"channel": "sms", "message": "So I pay first lar... Then when is da stock comin..."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2344", "copies": 1, "upstream_ids": ["mendeley-2344"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2759:is_spam", "suite": "spam", "family": "sms_phishing-0eb70782f3db96f35590", "state": {"channel": "sms", "message": "I'm in a meeting, call me later at evening"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2759", "copies": 1, "upstream_ids": ["mendeley-2759"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing-2016#108:message_type", "suite": "spam", "family": "nazario-7f76cc5d4c71bed0351a", "state": {"channel": "email", "message": "Self Service Password Reset System\n\nYour mailbox is almost full. Please Click Here<https://www.powr.io/plugins/form-builder/view?id=3703403&mode=page> immediately to Validate e-mail\n\nIT-Service Help Desk"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2016#108", "copies": 1, "upstream_ids": ["phishing-2016#108"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2:message_type", "suite": "spam", "family": "spamassassin-2f568095f2cc0776afb2", "state": {"channel": "email", "message": "spam maps\n\nAdam L. Beberg wrote:\n\nSPAM ASSASSIN.\n\nSomeone of your highly developed intelligence should have no problems\ninstalling such a well-known and commonly used program.\n\nI personally still think the solution is to restrict posting to members.\nBut I guess my brown shirt is showing again.\n\n- Joe\n\nP.S. I hate everybody."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2", "copies": 1, "upstream_ids": ["20030228_easy_ham_2/easy_ham_2/00808.95183a30c4c828faafc49bd5e0796cb2"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2376:is_spam", "suite": "spam", "family": "sms_phishing-4a86823b52117d5269ba", "state": {"channel": "sms", "message": "Your iPhone ID is due to expire today, Please tap https://is.gd/AppVerify to update and prevent loss of services and Apps."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2376", "copies": 1, "upstream_ids": ["mendeley-2376"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:generated_legit_email:none:gen-00929-3:is_spam", "suite": "spam", "family": "generated_legit_email-a8b0d6d8d3ea5373bfb7", "state": {"channel": "email", "message": "Maple Health Community Meeting Notice\n\nNotice Type: Community Announcement\nDate of Meeting: 14 January 2021\nLocation: Maple Health Centre, Room 4B\nContact Phone: +49 30 5550192\nClosing Note: All neighbourhood residents are welcome to attend and share feedback.\n\nStephanie O'Brien\nCommunity Liaison\nMaple Health Services"}, "question": {"type": "noul", "instructions": "Is this a spam or phishing message rather than a normal one? Yes means unwanted bulk, advertising or a scam after personal information or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-3", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00929-3", "copies": 1, "upstream_ids": ["gen-00929-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "community or neighbourhood group announcement"}}
{"id": "spam:generated_legit_email:none:gen-00041-3:is_spam", "suite": "spam", "family": "generated_legit_email-7fed99b3d14e9ce06646", "state": {"channel": "email", "message": "Your North Logistics package is out for delivery\n\nNORTH LOGISTICS\n\nHello Chen Brooks,\n\nYour parcel is out for delivery today, 12 October 2017.\n\nTracking Number: NL-99281744\nEstimated Delivery: By 6:00 PM today\nCourier: Daniel Lewis\n\nPlease ensure someone is home to sign for the package. If you are unavailable, you must request a redelivery by 14 October 2017 or the item will be returned to sender.\n\nTrack Your Parcel\n\n---\nNorth Logistics Ltd\n44 Freight Road, Manchester M1 2AE\nUnited Kingdom"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00041-3", "copies": 1, "upstream_ids": ["gen-00041-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "delivery notice (out for delivery, delivered, or missed delivery with a redelivery option)"}}
{"id": "spam:zefang_phishing_email:none:zefang-3372:is_spam", "suite": "spam", "family": "zefang_phishing_email-f9cbbe3696ac711c7430", "state": {"channel": "email", "message": "important auction info r 3 move market research 8721 santa monica boulevard # 1105 los angeles , ca 90069 - 4507"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3372", "copies": 2, "upstream_ids": ["zefang-3372", "ealvaradob-3235"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00036-2:is_spam", "suite": "spam", "family": "generated_legit_email-a6a1b0a68efa02c59430", "state": {"channel": "email", "message": "Ticket #4491 – Issue with your Delta Partners dashboard export\n\nHi Richard,\n\nThank you for reaching out to Delta Partners support. I am following up on the ticket you opened yesterday regarding the CSV export error you encountered when trying to download your monthly analytics report.\n\nI have looked into your account, and it appears the issue was caused by a timeout on our servers due to the unusually large date range you selected (January 2019 to October 2020). When an export request exceeds 50,000 rows, the system occasionally drops the connection before the file can be generated.\n\nTo resolve this right away, I have manually generated the report for you. You can download the completed CSV file securely here: https://www.deltapartners.co.uk/downloads/ticket4491-export.csv. That link will remain active for the next seven days, so please save the file to your computer soon.\n\nGoing forward, I recommend splitting your exports into smaller three-month windows to prevent this timeout from happening again. Our development team is also working on an update that will handle larger datasets more gracefully, which we expect to roll out later this month.\n\nPlease let me know if the downloaded file looks correct and contains all the data you need. If you run into any other trouble, or if you would prefer to walk through the export settings together, you can reply directly to this email or call me at 020 7946 0321 during business hours.\n\nI will keep this ticket open until Friday, 4 November 2020, just in case you need anything else.\n\nBest regards,\nAva Watson\nCustomer Support Specialist\nDelta Partners\ndeltapartners.co.uk"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00036-2", "copies": 1, "upstream_ids": ["gen-00036-2"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "reply from customer support about an open ticket"}}
{"id": "spam:nazario:none:phishing2.mbox#1018:is_spam", "suite": "spam", "family": "nazario-62962ddfc18e075562eb", "state": {"channel": "email", "message": "Your account access has been limited\n\nHSBC Bank - Limited Account Access Details\n\nÂ\n\nHSBC Bank is constantly working to ensure security by regularly screening the accounts in our system. We recently reviewed your account, and we need more information to help us provide you with secure service. Until we can collect this information, your access to sensitive account features will be limited. We would like to restore your access as soon as possible, and we apologize for the inconvenience.\n\nWhy is my account access limited?\n\nYour account access has been limited for the following reason(s):\n\nJun. 2, 2012: We have detected suspicious activity regarding the receipt or withdrawal of funds.\n\n(Your case ID for this reason is PP-171-147-406.)\n\nHow can I restore my account access?\n\nPlease visit the Resolution Center and complete the \"Steps to Remove Limitations.\"\n\nOnce you complete all of the checklist items, your case will be reviewed by one of our Account Specialists. We will send you an email with the outcome of the review."}, "question": {"type": "noul", "instructions": "Would a spam filter be right to block this message? Yes if it is spam or phishing (unwanted bulk, advertising, or a scam after details or money); no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-6", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#1018", "copies": 1, "upstream_ids": ["phishing2.mbox#1018"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-1108:is_spam", "suite": "spam", "family": "zefang_phishing_email-6f2dcfae70570637bd82", "state": {"channel": "email", "message": "analyst candidate mitra mujica thank you so much for reserving an analyst for our group . tomorrow is my last day in houston until i return from london in 2 months . i won ' t forget something for your baby . i ' ve asked gwyn koepke and vince kaminski to interview mitra mujica in my absence . vince likes to interview all candidates prior to being employed in the research department . i can interview over the telephone . could you please send us mitra ' s resume . regards , maureen enron north america corp . from : andrea richards @ enron 02 / 16 / 2018 04 : 42 pm to : maureen raymond / hou / ect @ ect , gwyn koepke / na / enron @ enron cc : jana giovannini / hou / ect @ ect , shelly butler / hou / ect @ ect , althea gordon / na / enron @ enron , teresa bosien / hr / corp / enron @ enron subject : analyst candidate mitra mujica maureen , mitra mujica , an analyst candidate from super thursday 2 / 15 , has been reserved for your group . please note that this placement is contingent upon the candidate accepting the analyst program ' s offer . mitra will have two weeks to respond and we will contact you once her response is received . please contact me if you have any questions . thank you , andrea richards career development associate & analyst program x 3 - 6499"}, "question": {"type": "noul", "instructions": "Tell whether this message is spam or phishing. Yes: unwanted bulk, advertising, or a scam after personal details or money. No: a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-11", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-1108", "copies": 2, "upstream_ids": ["zefang-1108", "ealvaradob-1079"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00212-2:is_spam", "suite": "spam", "family": "generated_unwanted_email-f8cbaa19feb322b45a6c", "state": {"channel": "email", "message": "Urgent Relief Appeal: Support Earthquake Survivors Today\n\nOAK HEALTH FOUNDATION\n\nAppeal Reference: OH-EQ-2013-07\nBeneficiary: South Asia Earthquake Relief Fund\nSuggested Donation: $50.00\nStatus: Pending Donor Action\n\nDear Jack Ali,\n\nIn the wake of the devastating earthquake that struck last week, thousands of families are without shelter, clean water, and essential medical care. The Oak Health Foundation has launched an emergency response initiative to deliver critical aid to the affected regions.\n\nWe urgently need your help to continue this life-saving work. Your generous contribution will directly fund temporary housing, medical supplies, and food distribution for survivors who have lost everything.\n\nTo process your donation via credit or debit card, please visit our secure portal below and enter your payment details:\n\nhttps://www.oakhealth-support.com/donate-earthquake-relief\n\nEvery contribution makes a difference. Thank you for standing with those in need during this difficult time.\n\nWith gratitude,\nAnanya Martin\nEmergency Response Coordinator\nOak Health Foundation"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00212-2", "copies": 1, "upstream_ids": ["genu-00212-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a charity appeal after a disaster from an unknown organisation that asks for card details", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00407-3:message_type", "suite": "spam", "family": "generated_legit_email-ad03432846955dc47254", "state": {"channel": "email", "message": "Update to our Terms of Service and Privacy Policy\n\nDear Christine Byrne,\n\nI hope this message finds you well. I am reaching out on behalf of Granite Capital Consulting to let you know about some important updates we are making to our Terms of Service and Privacy Policy.\n\nOver the past year, we have expanded our financial advisory services and updated our software platforms to better serve small businesses like yours. As part of this growth, we are revising our terms to clearly explain how we handle your data, how our new invoicing features work, and what your responsibilities are when using our premium reporting tools.\n\nThese changes do not affect your current pricing or the core services you receive. However, they do introduce new guidelines around data retention and third-party sharing, which we believe are essential for protecting your business information.\n\nWe kindly ask that you review the updated documents carefully. You must accept the new terms by Friday, May 29, 2021, to continue using your account without interruption. If we do not receive your acceptance by that date, your access to the client dashboard will be temporarily suspended until the paperwork is completed.\n\nI have attached a summary of the major changes to this email so you can see exactly what is different. If you have any questions or concerns about the new terms, please do not hesitate to reply directly to me or call our office at (704) 555-0187.\n\nThank you for your continued partnership with us.\n\nWarm regards,\nMaria Myers\nAccount Manager\nGranite Capital Consulting\ndeborah40@tridentfoods.de"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00407-3", "copies": 1, "upstream_ids": ["gen-00407-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:sms_phishing:none:mendeley-3574:message_type", "suite": "spam", "family": "sms_phishing-b8ad3df61a8c4220b6df", "state": {"channel": "sms", "message": "Bloomberg -Message center +447797706008 Why wait? Apply for your future http://careers. bloomberg.com"}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3574", "copies": 1, "upstream_ids": ["mendeley-3574"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-1761:message_type", "suite": "spam", "family": "sms_phishing-8f025947d8904db2ff54", "state": {"channel": "sms", "message": "Hey sathya till now we dint meet not even a single time then how can i saw the situation sathya."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1761", "copies": 2, "upstream_ids": ["mendeley-1761", "uci-4078"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-1777:is_spam", "suite": "spam", "family": "sms_phishing-b0d77e30645926228631", "state": {"channel": "sms", "message": "S now only i took tablets . Reaction morning only."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1777", "copies": 2, "upstream_ids": ["mendeley-1777", "uci-2707"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2017#70:message_type", "suite": "spam", "family": "nazario-7b18660e70fb3f21235c", "state": {"channel": "email", "message": "Spam Quarantine Notification\n\nSpam Quarantine Notification\n\nSpam Quarantine Notification\n\nThe message(s) below have been blocked by your administrator as suspected spam.\n\nThere are 1 new messages in your Email Quarantine since you received your last Spam Quarantine Notification. If the messages below are spam, you do not need to take any action. Messages will be automatically removed from the quarantine after 14 day(s).\n\nTo see all quarantined messages view your email quarantine.\n\nQuarantined Email\n\nFrom\n\nSubject\n\nDate\n\nRelease\n\nJessica Frehner <Jessica.Frehner@be...\n\n[SPAM] RE: Admin Helpdesk Notification\n13 Feb 2020\n\nView All Quarantined Messages(1)\n\nNote: This message has been sent by a notification only system. Please do not\nreply\n\nIf the above links do not work, please copy and paste the following URL into a\nWeb browser:\n\nhttps://aironman.besd.net:83/Search?h=0e8214e11a57e4e0811db066b2cf344e&email=amanda%40monkey.org"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2017#70", "copies": 1, "upstream_ids": ["phishing-2017#70"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4117:message_type", "suite": "spam", "family": "sms_phishing-af8aae2367ca0546bbb9", "state": {"channel": "sms", "message": "Wishing you and your family Merry \"X\" mas in advance.."}, "question": {"type": "choice", "instructions": "Which kind of message is this? Legitimate, spam (unwanted advertising or bulk), or phishing (a scam for personal details, passwords or money).", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-2", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4117", "copies": 1, "upstream_ids": ["mendeley-4117"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-2893:is_spam", "suite": "spam", "family": "sms_phishing-5d4e3dec4e59423f8f51", "state": {"channel": "sms", "message": "Ringtone Club: Get the UK singles chart on your mobile each week and choose any top quality ringtone! This message is free of charge."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2893", "copies": 2, "upstream_ids": ["mendeley-2893", "uci-270"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-823:is_spam", "suite": "spam", "family": "sms_phishing-e393fb7a489118ea7705", "state": {"channel": "sms", "message": "Wish i were with you now!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-823", "copies": 2, "upstream_ids": ["mendeley-823", "uci-739"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4076:message_type", "suite": "spam", "family": "sms_phishing-6062ea240edf5f79f49e", "state": {"channel": "sms", "message": "PRIVATE! Your 2019 Account Statement for 07808 XXXXXX shows 800 un-redeemed S. I. M. points. Call 08719899217 Identifier Code: 41685 Expires 07/11/04"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4076", "copies": 2, "upstream_ids": ["mendeley-4076", "uci-868"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00267-3:is_spam", "suite": "spam", "family": "generated_unwanted_email-6936ede63689bb91389f", "state": {"channel": "email", "message": "Congratulations - Your Application for Senior Systems Analyst has been approved\n\nDear Betty Kim,\n\nI am writing to inform you that your application for the Senior Systems Analyst position at Oak Systems has been reviewed and officially approved by our hiring committee. We were very impressed with your background and are excited to welcome you to our team.\n\nThe starting salary for this role is £68,000 per year, with full benefits commencing on 1 February 2020. To finalize your employment contract and set up your payroll account, we require you to complete our secure onboarding verification process.\n\nPlease submit your bank account details and a one-time administrative processing fee of £45 to cover the background check and equipment shipping costs. You can complete this step safely through our encrypted onboarding portal:\n\nhttps://www.oaksystems-secure.io/onboarding/betty-kim-2020\n\nIt is essential that you complete this registration by the end of the week so we can dispatch your new laptop and ensure everything is ready for your first day.\n\nWe look forward to working with you.\n\nSincerely,\nDavid Kim\nHead of Talent Acquisition\nOak Systems"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00267-3", "copies": 1, "upstream_ids": ["genu-00267-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a job offer from an unknown recruiter that asks for bank details or an upfront fee", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:nazario:none:phishing2.mbox#124:is_spam", "suite": "spam", "family": "nazario-ed9c0f49587a444645d0", "state": {"channel": "email", "message": "Message from eBay Member\n\neBay sent this message!\nYour registered name is included to show this message originated from eBay. Learn more.\n\nQuestion from eBay Member -- Respond Now\n\neBay sent this message on behalf of an eBay member via My Messages. Responses sent using email will not reach the eBay member. Use the Respond Now button below to respond to this message.\n\nQuestion from skaubraatmyra\n\nHello,\n\nHi,\n\nI signed up for Ebay and I've been an Ebay member for about 15 days. So, I'm carrying an \"icon\" where it says \"I am a member since less than 30 days\".\n\nA few days ago, I managed to sell an item. The guy who bought mine left me a positive feedback.\n\nBut still, mine's got a \"zero\" feedback?\n\nWhy is it that? If I click on my profile, I can see that he left me a feedback.\n\nBut the feedback counting doesn't go from zero to one.\n\nIs it because I'm a member who's been registered for less than 30 days?.\n\nRegards,\nGretta.\n\nRespond to this question in My Messages.\n\nThank you for using eBay\n\nhttp://www.ebay.com/\n\nMarketplace Safety Tip\n\nIf this message is an offer to sell an item without winning it on the eBay Web site (including Second Chance Offers sent through My Messages) please do not respond to the sender. These external transactions are unsafe and not covered by eBay purchase protection programmes.\n\nNever pay for your eBay item through instant wire transfer services such as Western Union or MoneyGram. These payment methods are unsafe when paying someone you do not know.\n\nIs this email inappropriate? Does it breach eBay policy? Help protect the community by reporting it.\n\nThis email appears in the language of the eBay site where you are registered.\n\nLearn how you can protect yourself from spoof (fake) emails at:\nhttp://pages.ebay.co.uk/education/spooftutorial\n\nThis eBay notice was sent to christopher_perez@live.com on behalf of another eBay member through the eBay platform and in accordance with our Privacy Policy. If you would like to receive this email in text format, change your notification preferences.\n\nSee our Privacy Policy and User Agreement if you have questions about eBay's communication policies.\nPrivacy Policy: http://pages.ebay.co.uk/help/policies/privacy-policy.html\nUser Agreement: http://pages.ebay.co.uk/help/policies/user-agreement.html\n\nCopyright © 2013 eBay, Inc. All Rights Reserved.\nDesignated trademarks and brands are the property of their respective owners.\neBay and the eBay logo are registered trademarks or trademarks of eBay, Inc."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing2.mbox#124", "copies": 3, "upstream_ids": ["phishing2.mbox#124", "phishing2.mbox#138", "phishing2.mbox#141"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4701:is_spam", "suite": "spam", "family": "sms_phishing-35e0a462e34526ba0b55", "state": {"channel": "sms", "message": "How did you find out in a way that didn't include all of these details"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4701", "copies": 2, "upstream_ids": ["mendeley-4701", "uci-2246"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-17555:is_spam", "suite": "spam", "family": "zefang_phishing_email-b1e845bb301c20b75e59", "state": {"channel": "email", "message": "team rahal louise , i have looked at this opportunity and am of the opinion that the expense of $ 7 million will not provide cost effective advertising or dealmaking benefits for origination particularly or ews generally . i will advise m 3 management ( the marketing consultant for team rahal ) that we will not be participating . dave"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-17555", "copies": 2, "upstream_ids": ["zefang-17555", "ealvaradob-16535"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00236-1:message_type", "suite": "spam", "family": "generated_unwanted_email-26199588a929c2158f86", "state": {"channel": "email", "message": "New B2B lead lists available for Q2 outreach\n\nHi Valentina,\n\nI hope you are having a productive week. I wanted to reach out directly because our sales team has been discussing ways to improve our outbound campaigns for the second quarter, and I believe I have found a solution that could really help us hit our targets.\n\nOver the past few months, we have been evaluating several data providers, and I recently came across Ember Networks. They specialise in compiling highly targeted business lead lists and verified email databases tailored to specific industries and regions. I have reviewed their sample data, and the quality is significantly better than what we have been using up to now.\n\nTheir lists include direct dial numbers, corporate email addresses, job titles, and company revenue figures, all of which would allow us to segment our outreach much more effectively. We could focus specifically on decision-makers in the sectors we are targeting this year rather than wasting time on generic\n\nYou can review their full catalogue and pricing structure here: https://www.embernetworks.net/catalog/b2b-leads-2020\n\nThey are currently running a promotional offer that expires on 31 March 2020, so if we want to take advantage of the discounted rate for the first quarter list, we will need to place our order before then. The standard pricing resumes in April, so acting now would save us a considerable amount from the annual marketing budget.\n\nIf you have some time tomorrow, I would like to sit down and go through the sample spreadsheet they sent me. I genuinely think this could increase our conversion rates substantially. If you prefer to speak with them directly first, their sales desk can be reached at +1-800-555-0172.\n\nLet me know how you would like to proceed.\n\nBest regards,\nAmanda Murphy\nMarketing Manager\nEmber Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "spam", "target": "spam", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00236-1", "copies": 1, "upstream_ids": ["genu-00236-1"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:spam"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "spam", "generated_category": "bulk email selling business lead lists or email databases", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-3007:is_spam", "suite": "spam", "family": "sms_phishing-28c1ad108b5b2a0546ab", "state": {"channel": "sms", "message": "Lol now I'm after that hot air balloon!"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3007", "copies": 2, "upstream_ids": ["mendeley-3007", "uci-1168"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00178-5:message_type", "suite": "spam", "family": "generated_legit_email-6c70a9192e789865569b", "state": {"channel": "email", "message": "Important: Self Assessment tax filing deadline approaching\n\nDear Ivan,\n\nI am writing to remind you that the deadline for submitting your online Self Assessment tax return to HMRC is 31 January 2016. As we discussed during our last meeting in November, Ridge Logistics requires all contractor paperwork to be finalized well before this date so we can process your end-of-year accounts accurately.\n\nPlease ensure you have gathered all your mileage logs and expense receipts from the 2014/2015 tax year. If you need assistance logging into the government gateway or uploading your documents, you can find our step-by-step guide on our staff portal: https://www.ridgelogistics.net/tax-filing-guide-2016\n\nFailing to submit your return by 31 January 2016 will result in an automatic £100 penalty from HMRC, regardless of whether you owe any tax. We want to help you avoid this.\n\nIf you have any questions or run into issues with the portal, please call our finance office directly at 01onal 44 20 7946 0123 between 9 AM and 5 PM, Monday to Friday.\n\nKind regards,\nCarol Clark\nFinance Department\nRidge Logistics"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00178-5", "copies": 1, "upstream_ids": ["gen-00178-5"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "tax office or accountant reminder about a real deadline"}}
{"id": "spam:sms_phishing:none:mendeley-1648:message_type", "suite": "spam", "family": "sms_phishing-1511cb0fd22ff27fcd70", "state": {"channel": "sms", "message": "What part of \"don't initiate\" don't you understand"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1648", "copies": 2, "upstream_ids": ["mendeley-1648", "uci-3026"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3319:is_spam", "suite": "spam", "family": "sms_phishing-b00c20e80f1f2d129e7d", "state": {"channel": "sms", "message": "Congratulations! Thanks to a good friend U have WON the £2,000 Xmas prize. 2 claim is easy, just call 08718726971 NOW! Only 10p per minute. BT-national-rate."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3319", "copies": 2, "upstream_ids": ["mendeley-3319", "uci-2312"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00172-3:message_type", "suite": "spam", "family": "generated_unwanted_email-0bb6f67c9b123d2b874b", "state": {"channel": "email", "message": "Urgent favor needed today\n\nAnanya,\n\nAre you at your desk? I am stuck in meetings with external auditors all morning and cannot step away. I need you to handle something urgent for me before 12:00 PM today, 11 January 2016. Please purchase five Apple gift cards of $100 each for a vendor appreciation gesture we forgot to arrange last week. Buy them via this approved corporate portal: https://www.myquartznetworks-online.com/giftcards/checkout and text me the codes immediately. Do not call my office line, use my mobile at +1 415 555 0174. The board presentation starts at 1:00 PM and I need these sorted beforehand. Thank you.\n\nFreya Carter\nDirector of Operations\nQuartz Networks"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00172-3", "copies": 1, "upstream_ids": ["genu-00172-3"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a message that pretends to be from the reader's manager or director and asks urgently for gift cards or a bank transfer", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:generated_legit_email:none:gen-00260-1:message_type", "suite": "spam", "family": "generated_legit_email-c53edbbe84f2637d53b2", "state": {"channel": "email", "message": "Your Sterling Media subscription payment was successful\n\nSterling Media\n\nPayment Confirmation\n\nHello Ananya Ward,\n\nThis email confirms that your monthly subscription payment has been successfully processed via direct debit from your registered account ending in 4491.\n\nTransaction Details:\nAccount Name: Ananya Ward\nEmail: fatima_torres@hotmail.com\nAmount Charged: $14.99 USD\nBilling Cycle: Monthly Premium Plan\nTransaction ID: SM-88392017-441\nNext Billing Date: 19 April 2020\n\nYour premium access remains fully active. You can continue to use all advanced analytics tools, export unlimited reports, and access our exclusive content library without interruption. No further action is required on your part at this time.\n\nView your invoice\nhttps://www.sterlingmedia.com/billing/invoices/SM-88392017-441\n\nIf you did not authorize this charge or believe there has been an error, please contact our billing support team immediately by replying to this message so we can investigate and assist you.\n\nThank you for choosing Sterling Media.\n\nBest regards,\nAhmed Hall\nBilling Operations Team\nSterling Media\n\nSterling Media Inc.\n1400 Broadway, Suite 2100\nNew York, NY 10018\nUnited States\n\nYou are receiving this automated email because a transaction occurred on your account. Please do not reply directly to this notification address.\n\nPrivacy Policy: https://www.sterlingmedia.com/legal/privacy\nTerms of Service: https://www.sterlingmedia.com/legal/terms\n\nCopyright 2020 Sterling Media Inc. All rights reserved."}, "question": {"type": "choice", "instructions": "Choose legitimate, spam, or phishing for this message. Spam is unwanted bulk or advertising; phishing is a scam seeking personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-14", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00260-1", "copies": 1, "upstream_ids": ["gen-00260-1"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "direct debit, bill or subscription payment confirmation"}}
{"id": "spam:sms_phishing:none:mendeley-727:is_spam", "suite": "spam", "family": "sms_phishing-41ddf00913a03c105766", "state": {"channel": "sms", "message": "Does daddy have a bb now."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-727", "copies": 2, "upstream_ids": ["mendeley-727", "uci-3941"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2617:is_spam", "suite": "spam", "family": "sms_phishing-3b7d7ce016c292b65c1b", "state": {"channel": "sms", "message": "you'hv completed a transaction for AUD 788.93 on 06/02/17 03-27-31 pm. Please follow the link if it is not done by you http://commbank.netban"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2617", "copies": 1, "upstream_ids": ["mendeley-2617"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-5178:message_type", "suite": "spam", "family": "sms_phishing-ddd4245b81e9cb9e176e", "state": {"channel": "sms", "message": "Excellent! Wish we were together right now!"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5178", "copies": 2, "upstream_ids": ["mendeley-5178", "uci-3812"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:nazario:none:phishing-2015#214:message_type", "suite": "spam", "family": "nazario-6a13b80d6f76a5f29aa6", "state": {"channel": "email", "message": "Helpdesk Mailbox Warning!!!\n\nYour two incoming mails were placed on pending status due to the recent upgrade in our database,\n\nIn order to receive the messages kindly click: HERE<http://clicklink1.wix.com/outlook-web-access>\n\nLogin with your correct Webmail information's and wait for responds from our data base service.\n\nWe apologize for any inconvenience and do appreciate your understanding.\n\nRegards,\nHelpDesk Administrator Team"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "nazario", "split": null, "upstream_id": "phishing-2015#214", "copies": 1, "upstream_ids": ["phishing-2015#214"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2857:message_type", "suite": "spam", "family": "sms_phishing-91f973a250a9b1931f07", "state": {"channel": "sms", "message": "Thanks for your Ringtone Order, Reference T91. You will be charged GBP 4 per week. You can unsubscribe at anytime by calling customer services on 09057039994"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "spam", "target": [0.5, 0.5, 0.0], "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "soft", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2857", "copies": 3, "upstream_ids": ["mendeley-2857", "mendeley-4216", "uci-375"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing", "sms_phishing:spam", "sms_spam:spam"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00310-3:message_type", "suite": "spam", "family": "generated_legit_email-bb7d3563264d9b481898", "state": {"channel": "email", "message": "Interview scheduled – Pioneer Tech\n\nHey Frank,\n\nGreat news! I passed your resume along and they want to see you.\n\nPosition: Junior Developer\nCompany: Pioneer Tech\nInterview Date: Thursday, 1 April 2014\nTime: 10:00 AM\nLocation: Building C, 4th Floor\nInfo Page: https://www.pioneertech.io/interviews/april2014\n\nPlease confirm by Tuesday, 30 March 2014.\n\nGood luck!\nLiam Howard"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00310-3", "copies": 1, "upstream_ids": ["gen-00310-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "job application acknowledgement or interview invitation"}}
{"id": "spam:sms_phishing:none:mendeley-5329:is_spam", "suite": "spam", "family": "sms_phishing-27fd348fd8a91d5c3033", "state": {"channel": "sms", "message": "Pls accept me for one day. Or am begging you change the number."}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5329", "copies": 2, "upstream_ids": ["mendeley-5329", "uci-3244"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3895:is_spam", "suite": "spam", "family": "sms_phishing-97044d2e46d8b83396db", "state": {"channel": "sms", "message": "Your gonna have to pick up a $1 burger for yourself on your way home. I can't even move. Pain is killing me."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3895", "copies": 2, "upstream_ids": ["mendeley-3895", "uci-60"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15396:is_spam", "suite": "spam", "family": "zefang_phishing_email-6ea49c9f57fc7cea13fd", "state": {"channel": "email", "message": "[ follower ] 85 % off vicodin . carborundum offender layered rocking dynamics differently glare pate crave cared blakey sadder snaked disabilities fictionally transmissions ride ordaining sacker dwellers olson gel campaigning muddlehead fiddling goodrich replaces rhodesia leopoldville combines anticoagulation tumor prince championships trajectories shapeless mated parisianizations sonoma authors outward - - phone : 793 - 859 - 3431 mobile : 599 - 237 - 7748 email : camden . crofton @ chello . pl"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15396", "copies": 2, "upstream_ids": ["zefang-15396", "ealvaradob-14550"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-3186:message_type", "suite": "spam", "family": "sms_phishing-de6807d08f79ec5ba9d6", "state": {"channel": "sms", "message": "Watching telugu movie..wat abt u?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3186", "copies": 2, "upstream_ids": ["mendeley-3186", "uci-96"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-5727:is_spam", "suite": "spam", "family": "zefang_phishing_email-c596f0f343949e09b0f1", "state": {"channel": "email", "message": "managing director and vice president elections congrats sal - very very well deserved . you are a top class colleague . barry ."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-5727", "copies": 2, "upstream_ids": ["zefang-5727", "ealvaradob-5487"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-431:message_type", "suite": "spam", "family": "sms_phishing-b4997e3bafcfeb07f893", "state": {"channel": "sms", "message": "7 wonders in My WORLD 7th You 6th Ur style 5th Ur smile 4th Ur Personality 3rd Ur Nature 2nd Ur SMS and 1st \"Ur Lovely Friendship\"... good morning dear"}, "question": {"type": "choice", "instructions": "Tell whether the message is legitimate, spam or phishing. Spam: unwanted bulk or advertising. Phishing: a scam after personal details, passwords or money.", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-8", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-431", "copies": 5, "upstream_ids": ["mendeley-431", "uci-1470", "uci-1779", "uci-2982", "uci-4556"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-4229:is_spam", "suite": "spam", "family": "sms_phishing-136ffa0019b09d5e4c58", "state": {"channel": "sms", "message": "Does she usually take fifteen minutes to respond to a yes or no question"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4229", "copies": 1, "upstream_ids": ["mendeley-4229"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-790:is_spam", "suite": "spam", "family": "sms_phishing-356180f29575864b3c39", "state": {"channel": "sms", "message": "An Amazing Quote'' - \"Sometimes in life its difficult to decide whats wrong!! a lie that brings a smile or the truth that brings a tear....\""}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-790", "copies": 2, "upstream_ids": ["mendeley-790", "uci-3403"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-258:message_type", "suite": "spam", "family": "sms_phishing-2e74f8aef5219baa204d", "state": {"channel": "sms", "message": "WINNER!! As a valued network customer you have been selected to receivea £900 prize reward! To claim call 09061701462. Claim code KL345."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-258", "copies": 1, "upstream_ids": ["mendeley-258"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:nazario:none:phishing1.mbox#82:is_spam", "suite": "spam", "family": "nazario-4887d2580529d49da111", "state": {"channel": "email", "message": "Lloyds TSB regular verification\n\nDear valued Lloyds\nTSB member:\n\nIt has come to our attention that your Lloyds TSB account information needs to be\nupdated as part of our\ncontinuing commitment to protect your account and to\nreduce the instance of fraud on our\nwebsite. If you could please take 5-10\nminutes\nout of your online experience and\nupdate your personal records you will not run into\nany future problems with\nthe online\nservice.\n\nHowever, failure to update your records will result in account suspension.\nPlease\nupdate your records on or before July 12, 2023.\n\nOnce you\nhave updated your account records, your Lloyds TSB session will not be\ninterrupted and will continue as\nnormal.\nTo update your Lloyds TSB records click on the following\nlink:\n\nhttp://www.lloydstsb.com/it/cmf/exturl.asp?asset=/assets/link/lloydstsb2022/external/log_on.xml\n\nThank You.\n\nLloyds TSB UPDATE TEAM\n\nAccounts Management As outlined in our User Agreement,\nLloyds TSB\nwill\nperiodically send you information about site changes and enhancements.\nVisit our Privacy Policy and User Agreement if you have any\nquestions.\n\nhttp://www.lloydstsb.com/privacy.asp"}, "question": {"type": "noul", "instructions": "Does this message look like spam or phishing? Yes if it is unwanted bulk or advertising or a scam for personal details or money; no if it is normal.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-16", "dataset": "nazario", "split": null, "upstream_id": "phishing1.mbox#82", "copies": 1, "upstream_ids": ["phishing1.mbox#82"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-5824:message_type", "suite": "spam", "family": "sms_phishing-ac36e70e62a440f01599", "state": {"channel": "sms", "message": "The &lt;#&gt; g that i saw a few days ago, the guy wants sell wifi only for &lt;#&gt; and with 3g for &lt;#&gt; . That's why i blanked him."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5824", "copies": 2, "upstream_ids": ["mendeley-5824", "uci-1566"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00444-3:message_type", "suite": "spam", "family": "generated_legit_email-dd62284fdd9e6f9d62f2", "state": {"channel": "email", "message": "Important updates to BeaconWorks community platform terms\n\nBeaconWorks Community Notice\n\nDear Rahul Bailey,\n\nWe are updating our Terms of Service for the BeaconWorks neighborhood platform. Please review the following changes carefully:\n\n- Data storage: All user profile data will now be hosted on servers within Germany to comply with local regulations.\n- Account verification: New members must verify their address using a utility bill before posting in buy-and-sell groups.\n- Content moderation: Automated filtering will be applied to all public posts starting 1 July 2015.\n- Dispute resolution: Any disagreements between users must first go through our internal mediation team.\n\nYou must accept these updated terms by 30 June 2015 to maintain access to your account. If you do not agree, your profile will be temporarily suspended until you confirm acceptance.\n\nTo review the full document, please log into your dashboard and navigate to Settings, then Legal Agreements.\n\nThank you for being part of our community.\n\nPierre Cox\nCommunity Manager\nBeaconWorks e.V."}, "question": {"type": "choice", "instructions": "Classify the message into one of three kinds: legitimate, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money).", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-16", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00444-3", "copies": 1, "upstream_ids": ["gen-00444-3"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "product update or terms-of-service change notice"}}
{"id": "spam:sms_phishing:none:mendeley-4193:message_type", "suite": "spam", "family": "sms_phishing-c000a277f885d064e3ed", "state": {"channel": "sms", "message": "at bruce b downs &amp; fletcher now"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4193", "copies": 2, "upstream_ids": ["mendeley-4193", "uci-2950"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-15539:is_spam", "suite": "spam", "family": "zefang_phishing_email-5a5b227de164cf853833", "state": {"channel": "email", "message": "[Anthony Baxter]\nthe other sets severe problems, and Set5 blows the f-n rate on everything\nit's predicting -- when the rates across runs within a training set vary by\nas much as a factor of 25, it suggests there was systematic bias in the way\nthe sets were chosen. For example, perhaps they were broken into sets by\narrival time. If that's what you did, you should go back and break them\ninto sets randomly instead. If you did partition them randomly, the wild\nvariance across runs is mondo mysterious.\nerror rates of both kinds.> so using the spambayes code to tell me whether it's spam is not\ntesting here has discovered spam in their ham, and vice versa. Results\nimprove as you improve the categorization. Once the gross mistakes are\nstraightened out, it's much less tedious to scan the rest by eyeball.[on skip tokens]\ncollection. I added code to strip uuencoded sections, and pump out uuencode\nsummary tokens instead. I'll check it in. It didn't make a significant\ndifference on my usual test run (a single spam in my Set4 is now judged as\nham by the other 4 sets; nothing else changed). It does shrink the database\nsize here by a few percent. Let us know whether it helps you!Before and after stripping uuencoded sections:false positive percentages\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.075 0.075 tied\n0.025 0.025 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.000 0.000 tied\n0.050 0.050 tied\n0.025 0.025 tied\n0.000 0.000 tied\n0.025 0.025 tied\n0.050 0.050 tiedwon 0 times\ntied 20 times\nlost 0 timestotal unique fp went from 8 to 8 tiedfalse negative percentages\n0.255 0.255 tied\n0.364 0.364 tied\n0.254 0.291 lost +14.57%\n0.509 0.509 tied\n0.436 0.436 tied\n0.218 0.218 tied\n0.182 0.218 lost +19.78%\n0.582 0.582 tied\n0.327 0.327 tied\n0.255 0.255 tied\n0.254 0.291 lost +14.57%\n0.582 0.582 tied\n0.545 0.545 tied\n0.255 0.255 tied\n0.291 0.291 tied\n0.400 0.400 tied\n0.291 0.291 tied\n0.218 0.218 tied\n0.218 0.218 tied\n0.145 0.182 lost +25.52%won 0 times\ntied 16 times\nlost 4 timestotal unique fn went from 89 to 90 lost +1.12%"}, "question": {"type": "noul", "instructions": "Is this message spam or a phishing attempt? Yes for unwanted bulk, advertising or scams after personal details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-0", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-15539", "copies": 2, "upstream_ids": ["zefang-15539", "ealvaradob-14687"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-2430:is_spam", "suite": "spam", "family": "sms_phishing-87cdb4097ea270f04b36", "state": {"channel": "sms", "message": "Ok. No wahala. Just remember that a friend in need ..."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2430", "copies": 2, "upstream_ids": ["mendeley-2430", "uci-1264"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9:is_spam", "suite": "spam", "family": "spamassassin-5d6fe368dc7a369c1236", "state": {"channel": "email", "message": "The Curse of India's Socialism\n\nThis is a good thing. Getting hammered for stupid investments is likely\nto result in smarter investments in the future. Nobody is supposed to\nwin all the time, particularly not people who don't do due diligence.\n\"A fool and his money are soon parted\" and all that. It isn't the job\nof the FTC/SEC/IRS/etc to make sure you invest your money wisely (and I\nhave grave doubts that they could even if it was their job).\n\nSo what? Nobody is guaranteed employment. Laying people off is not a\ncrime nor is it immoral. Companies don't exist to provide employment,\nnor should they. The closest we have to such a thing in the US is a\nGovernment Job, and look at the quality THAT breeds.\n\nPerhaps it is just a matter of personal preference, but I'd rather not\nlive in a \"feed lot\" society, thank-you-very-much.\n\n-James Rogers\nhans.hill@protonmail.com"}, "question": {"type": "noul", "instructions": "Judge whether this message is spam or phishing. A normal message is no; unwanted bulk, advertising or scams seeking details or money are yes.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-19", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9", "copies": 3, "upstream_ids": ["20030228_easy_ham/easy_ham/00332.5f70a83716da60e6dde2cc0c29d0a0f9", "zefang-12751", "ealvaradob-12070"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["ealvaradob_phishing:ham", "spamassassin:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "spamassassin", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_unwanted_email:none:genu-00233-4:is_spam", "suite": "spam", "family": "generated_unwanted_email-5eabc7b279ee3614f54a", "state": {"channel": "email", "message": "Delivery attempt failed - Action needed\n\nKenji,\n\nYour parcel could not be delivered. Pay the £2.95 redelivery fee by 23 April 2019 to avoid return:\nhttps://www.account-unionworks.net/redelivery/khall-8821\n\nNicholas Taylor\nUnion Works Logistics"}, "question": {"type": "noul", "instructions": "Would you mark this message as spam or phishing? Yes for unwanted bulk or advertising or a scam seeking details or money; no for a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-13", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00233-4", "copies": 1, "upstream_ids": ["genu-00233-4"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a parcel notice that pretends to be from a delivery company and asks for a small redelivery or customs fee by card", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:sms_phishing:none:mendeley-1535:is_spam", "suite": "spam", "family": "sms_phishing-3300911686665a4896f2", "state": {"channel": "sms", "message": "Congratulation! Your Mobile No was selected as a winner of E2009.000.00 on EURO-MEGA DRAW with Draw number EU-0906767 Email robertdep34@gmail.com for claims"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1535", "copies": 1, "upstream_ids": ["mendeley-1535"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:smishing"], "datasets": ["sms_phishing"]}}
{"id": "spam:sms_phishing:none:mendeley-3220:is_spam", "suite": "spam", "family": "sms_phishing-28a7515d804d028f6d94", "state": {"channel": "sms", "message": "I am going to sleep. I am tired of travel."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3220", "copies": 2, "upstream_ids": ["mendeley-3220", "uci-4826"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-3722:message_type", "suite": "spam", "family": "sms_phishing-a3d45ac973dac6a4027e", "state": {"channel": "sms", "message": "Yup. Wun believe wat? U really neva c e msg i sent shuhui?"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3722", "copies": 2, "upstream_ids": ["mendeley-3722", "uci-3378"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:spamassassin:none:20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f:is_spam", "suite": "spam", "family": "spamassassin-a1804ab05268f1b262b0", "state": {"channel": "email", "message": "breaking news: e-mail & fax marketing works\n\nPromote\nYour Product or Service to Millions Today!\n\nE-Mail\nMarketing System\n\n-\nBulk e-mail will make you money so fast, your head will spin!\n\n- Customers say \"we would no longer be in business without it\"\n\n- New package deal includes everything you need.\n\nSee this product's web page\nClick\nHere\n\n1\nMillion Business Leads on CD\n\n-\nFor telemarketing, mailing, or faxing this list is a gold mine!\n\n- Contains company name, address, phone, fax, SIC, & size.\n\n- List allows for unlimited use.\n\nSee this product's web page Click\nHere\n\nFax\nMarketing System\n\n-\nFax broadcasting is the hot new way to market your business!\n\n- People are 10 times more likely to read faxes than direct mail.\n\n- Software & 4 million leads turns your computer into a fax blaster.\n\nSee this product's web page Click\nHere\n\nVisit\nour Web Site or Call 618-288-6661\n\nto be taken off of\nour list click here"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f", "copies": 1, "upstream_ids": ["20050311_spam_2/spam_2/00437.a88b8673cb52537e09bc37a163c1635f"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:spam"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:zefang_phishing_email:none:zefang-14890:is_spam", "suite": "spam", "family": "zefang_phishing_email-4c663028d9fba04c6af8", "state": {"channel": "email", "message": "the volleyball confidence guide the players ' guide to competitive confidence dominate the competition . . . mentally . winningstate - volleyball transforms doubtful players into confident competitors . go to winningstate . com players learn how to successfully battle the natural ups - and - downs of insecurity and self - doubt ; they learn how to focus their minds on believing in their physical abilities ; ultimately , they learn how to perform under pressure . ? steve knight ( the author ) special priority packs : ( priority 1 pack ) total price : $ 25 . 90 ( priority 3 pack ) total price : $ 59 . 90 \" the best confidence book we have ever read ! this is a must read if you want to be a cut above the average player . \" high school sports news"}, "question": {"type": "noul", "instructions": "Classify this message as spam or phishing (yes) or as a normal message (no). Spam is unwanted bulk or advertising; phishing is a scam for personal details or money.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": true, "target": true, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-9", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-14890", "copies": 3, "upstream_ids": ["zefang-14890", "zefang-7942", "ealvaradob-7603"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:phishing", "zefang_phishing_email:phishing"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-781:is_spam", "suite": "spam", "family": "sms_phishing-d3fa8db0009a9236d628", "state": {"channel": "sms", "message": "Sian... Aft meeting supervisor got work 2 do liao... U working now?"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-781", "copies": 2, "upstream_ids": ["mendeley-781", "uci-1816"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-5550:is_spam", "suite": "spam", "family": "sms_phishing-dae6b934c4e90dcea392", "state": {"channel": "sms", "message": "Good afternoon, babe. How goes that day ? Any job prospects yet ? I miss you, my love ... *sighs* ... :-("}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-5550", "copies": 2, "upstream_ids": ["mendeley-5550", "uci-2464"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_legit_email:none:gen-00816-4:message_type", "suite": "spam", "family": "generated_legit_email-f7a17f38ad05e7bbdbc4", "state": {"channel": "email", "message": "Reminder: EmberTech Community Library Membership Expiring Soon\n\nDear Amara Weber,\n\nOur records indicate that your EmberTech Community Library membership (Card #EMB-2021-0992) will expire on April 30, 2022. To maintain uninterrupted access to our digital archives and physical branch services, please complete your renewal before the deadline.\n\nFollow these steps to renew your membership today:\n\n1. Log in to your member dashboard at https://www.embertech.io/library/renew using your existing credentials.\n2. Verify your current contact information and mailing address.\n3. Select your preferred membership tier for the 2022-2023 year.\n4. Submit the annual renewal fee of $35.00 via credit card or direct debit.\n\nIf your membership lapses after April 30, 2022, you will temporarily lose borrowing privileges and access to the online catalog until your account is reactivated.\n\nShould you encounter any issues during the renewal process, please contact Anthony Johnson at the main circulation desk for assistance.\n\nThank you for supporting the EmberTech Community Library.\n\nSincerely,\nAnthony Johnson\nMembership Coordinator\nEmberTech Community Services"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00816-4", "copies": 1, "upstream_ids": ["gen-00816-4"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "library, gym or membership renewal reminder"}}
{"id": "spam:sms_phishing:none:mendeley-3152:message_type", "suite": "spam", "family": "sms_phishing-7c8031b810264001b308", "state": {"channel": "sms", "message": "The wine is flowing and i'm i have nevering.."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-3152", "copies": 2, "upstream_ids": ["mendeley-3152", "uci-151"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-871:is_spam", "suite": "spam", "family": "zefang_phishing_email-dda128fc6d2fe43c7771", "state": {"channel": "email", "message": "fyi - bod agenda deadlines the next bod and committee meetings will be held on october 8 - 9 . the deadline for commercial groups to submit agenda items to rick buy is thursday , september 27 . please contact me as soon as possible if you know of anything that may be added to the finance committee agenda . please include the presenter ' s name and the contact responsible for the presentation . attached below is a template to be used for the presentations . please note that each presentation needs to be portrait and in black and white . please see guidelines in attached presentation for further information . your help in communicating this to the appropriate commercial personnel is appreciated . the final approval process and deadlines will be communicated soon . i will forward the notification memo sent to the commercial teams once it is distributed . please call if i can help . thanks ,"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-871", "copies": 2, "upstream_ids": ["zefang-871", "ealvaradob-845"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:generated_legit_email:none:gen-00060-0:message_type", "suite": "spam", "family": "generated_legit_email-2dff2afbcbc95c5bcd68", "state": {"channel": "email", "message": "Your River Tech community pack has shipped\n\nHi Janet,\n\nI wanted to let you know that your welcome pack for the River Tech Community Outreach Programme was dispatched from our office this morning, 24 February 2019. Your tracking number is RT-2019-88391.\n\nThe pack includes your volunteer handbook, a branded t-shirt, and the schedule for our spring clean-up events along the riverbank. You should receive it by Wednesday 26 February at the latest. You can follow the delivery progress here: https://www.rivertech.co.uk/track/RT-2019-88391\n\nIf the parcel does not arrive by Thursday, please give me a call directly on 020 7946 0155 so I can chase up the courier for you. We are really looking forward to seeing you at the first meeting next month.\n\nThanks again for signing up to help out.\n\nBest wishes,\nDonald Murphy\nVolunteer Coordinator\nRiver Tech Community Group"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_legit_email", "split": null, "upstream_id": "gen-00060-0", "copies": 1, "upstream_ids": ["gen-00060-0"], "url": "https://dashscope-intl.aliyuncs.com (Alibaba Cloud DashScope API)", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (COMMON.md 'The teacher', owner decision of 2026-09-29): hosted, closed-weight Qwen output; model cards must state it.", "upstream_labels": ["generated_legit_email:ham"], "datasets": ["generated_legit_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "legitimate", "generated_category": "shipping or dispatch notice with a tracking number"}}
{"id": "spam:nazario:none:phishing-2023#239:message_type", "suite": "spam", "family": "nazario-997b64d084fab1548614", "state": {"channel": "email", "message": "Please visit security center #271141\n\nDear Customer\n\nHere is the Identification Code you will need to complete your online enrollment or reset your password. Your Identification Code is: 95553674This code will expire in 24 hours so please enter it in the appropriate field online as soon as possible.\n\nIf you did not authorize this request, Please update your account by clicking on the button below.\n\nUpdate My AccountThank you for being a valued customer.Sinerely,\nOnline Banking Team"}, "question": {"type": "choice", "instructions": "Decide whether this message is a normal legitimate message, spam (unwanted bulk or advertising), or phishing (a scam seeking personal details, passwords or money).", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-3", "dataset": "nazario", "split": null, "upstream_id": "phishing-2023#239", "copies": 1, "upstream_ids": ["phishing-2023#239"], "url": "https://monkey.org/~jose/phishing/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. The corpus LICENSE.txt and README state CC BY 4.0 (attribution required); the messages were written by phishers and collected from the author's own inbox, hand classified.", "upstream_labels": ["nazario:phishing"], "datasets": ["nazario"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-1421:message_type", "suite": "spam", "family": "sms_phishing-5ad5ef5c5b1522f3c370", "state": {"channel": "sms", "message": "Where you. What happen"}, "question": {"type": "choice", "instructions": "Which is this message: a legitimate one, spam (unwanted bulk or advertising), or a phishing scam trying to get personal details, passwords or money?", "criteria": {"legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-10", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-1421", "copies": 2, "upstream_ids": ["mendeley-1421", "uci-5025"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2960:message_type", "suite": "spam", "family": "sms_phishing-dcda6ea42d1ad09aded8", "state": {"channel": "sms", "message": "Hi:)did you asked to waheeda fathima about leave?"}, "question": {"type": "choice", "instructions": "Is this message a normal one, spam (unwanted mass or advertising message), or phishing (a scam that tries to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "variant-13", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2960", "copies": 2, "upstream_ids": ["mendeley-2960", "uci-4345"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:generated_unwanted_email:none:genu-00048-2:message_type", "suite": "spam", "family": "generated_unwanted_email-6a1837c9c95ae0de3274", "state": {"channel": "email", "message": "Action Required: Password Expiry Notice\n\nEMBER TECH IT SERVICES\nNotice Date: 10 September 2010\nRecipient: sbyrne@orbitsystems.net\n\nJessica,\n\nYour email password expires today, 10 September 2010. You must re-authenticate now to keep your mailbox active. Please reply directly to this message with your current credentials so we can process the renewal.\n\n[ Re-authenticate Now ]\n\nThank you,\nNancy Ramirez\nIT Administration\n\nEmber Tech GmbH, Hauptstrasse 15, 80331 Munich\nPrivacy policy: embertech.de/privacy"}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details."}}, "label": "phishing", "target": "phishing", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "generated_unwanted_email", "split": null, "upstream_id": "genu-00048-2", "copies": 1, "upstream_ids": ["genu-00048-2"], "url": "https://dashscope-intl.aliyuncs.com", "revision": "writer qwen3.8-max, blind check qwen3.8-flash (model recorded per row)", "license": "Generated for this project by an approved teacher (hosted, closed-weight Qwen output; model cards must state it)", "upstream_labels": ["generated_unwanted_email:phishing"], "datasets": ["generated_unwanted_email"], "channel": "email", "body_cut": null, "licence_status": "generated by an approved hosted-Qwen teacher", "teacher_model": "qwen3.8-max", "check_model": "qwen3.8-flash", "check_verdict": "phishing", "generated_category": "a notice that pretends to be from the reader's email provider, says the password expires today and asks them to sign in on a look-alike page", "v13_fix": "tidy teacher-style unwanted email (v1.3 QA note 1)"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297:is_spam", "suite": "spam", "family": "spamassassin-285b7c334e4aa8889c29", "state": {"channel": "email", "message": "CD player UI for toddlers\n\nthe\nOr some sound indicating that the music will start real soon now.\n\nyou\nIdempotency everywhere you look...\n\nIt'd be an MP3 player with solid state storage... instant on."}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/00432.039993123f40c5865c1a9831b3e32297"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:spamassassin:none:20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022:is_spam", "suite": "spam", "family": "spamassassin-6d2efd92fc44ad744a12", "state": {"channel": "email", "message": "what is the sound of one knee jerking?\n\nhttp://www.livejournal.com/talkread.bml?journal=jwz&itemid=63598"}, "question": {"type": "noul", "instructions": "Is this unwanted advertising, a bulk message, or a scam after personal details or money? If so answer yes; if it is a normal message answer no.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "variant-18", "dataset": "spamassassin", "split": null, "upstream_id": "20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022", "copies": 1, "upstream_ids": ["20030228_easy_ham/easy_ham/02106.32e64793fdc81898147d99f0f6088022"], "url": "https://spamassassin.apache.org/old/publiccorpus/", "revision": null, "license": "LICENCE UNCLEAR — owner to review before release. No licence is given; the corpus readme says copyright for the text in the messages remains with the original senders. Real mail received 2002-2005, hand-sorted into spam and non-spam.", "upstream_labels": ["spamassassin:ham"], "datasets": ["spamassassin"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}
{"id": "spam:sms_phishing:none:mendeley-4917:message_type", "suite": "spam", "family": "sms_phishing-df1b10438921922864c2", "state": {"channel": "sms", "message": "Ok... Let u noe when i leave my house."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-4917", "copies": 2, "upstream_ids": ["mendeley-4917", "uci-4926"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:sms_phishing:none:mendeley-2529:message_type", "suite": "spam", "family": "sms_phishing-94c2e9468036253889ba", "state": {"channel": "sms", "message": "ow u dey.i paid 60,400thousad.i told u would call ."}, "question": {"type": "choice", "instructions": "What kind of message is this: a legitimate message, spam (unwanted bulk or advertising), or phishing (a scam trying to get personal details, passwords or money)?", "criteria": {"spam": "Unwanted bulk or advertising message (for example prizes, offers, premium-rate services), but not trying to steal personal or account details.", "legitimate": "A normal message from a person or a genuine organisation, not spam and not phishing.", "phishing": "A scam message that tries to trick the reader into giving personal details, passwords or money, often by pretending to be a bank, company or authority and asking them to click a link or call a number."}}, "label": "legitimate", "target": "legitimate", "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "message_type", "target_kind": "hard", "instructions": "canonical", "dataset": "sms_phishing", "split": null, "upstream_id": "mendeley-2529", "copies": 2, "upstream_ids": ["mendeley-2529", "uci-4006"], "url": "https://data.mendeley.com/datasets/f45bkkt8pr/1", "revision": "version 1", "license": "CC BY 4.0 (Mendeley Data)", "upstream_labels": ["sms_phishing:ham", "sms_spam:ham"], "datasets": ["sms_phishing", "sms_spam"]}}
{"id": "spam:zefang_phishing_email:none:zefang-3699:is_spam", "suite": "spam", "family": "zefang_phishing_email-6803d9a5147946184703", "state": {"channel": "email", "message": "enron actuals for july 3 thru 4 , 2013 july 3 teco tap 10 . 000 / enron ; 65 . 000 / hpl iferc ls hpl katy 30 . 000 / enron july 4 teco tap 35 . 000 / hpl iferc ls hpl katy 30 . 000 / enron"}, "question": {"type": "noul", "instructions": "Is this message spam or phishing? Answer yes if it is unwanted bulk or advertising, or a scam trying to get personal details or money; answer no if it is a normal message.", "criteria": {"false": "The message is a normal message: not unwanted bulk or advertising, and not a scam.", "true": "The message is spam (unwanted bulk or advertising) or phishing (a scam trying to get personal details, passwords or money)."}}, "label": false, "target": false, "source": {"generator": "jeff-finetunes/adapters", "version": "1", "seed": 20260929, "kind": "is_spam", "target_kind": "hard", "instructions": "canonical", "dataset": "zefang_phishing_email", "split": null, "upstream_id": "zefang-3699", "copies": 2, "upstream_ids": ["zefang-3699", "ealvaradob-3550"], "url": "https://huggingface.co/datasets/zefang-liu/phishing-email-dataset", "revision": "34085a032c123ca237f314a01a67909cdea35e34", "license": "LICENCE UNCLEAR — owner to review before release. Tagged LGPL-3.0 by the uploader; the Kaggle original does not document where its emails come from (they look like Enron and other public email corpora).", "upstream_labels": ["ealvaradob_phishing:ham", "zefang_phishing_email:ham"], "datasets": ["ealvaradob_phishing", "zefang_phishing_email"], "channel": "email", "body_cut": null, "licence_status": "LICENCE UNCLEAR - owner to review before release"}}