Add architecture diagram (assets/architecture.png + SVG source) and README Architecture section

.gitattributes

@@ -3,3 +3,4 @@
 *.pt filter=lfs diff=lfs merge=lfs -text
 tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 *.pdf filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -33,7 +33,7 @@ datasets:
 
 > **Licences — what covers what.**
 > - **Apache-2.0** ([LICENSE](LICENSE), [NOTICE](NOTICE)): the model weights (`model.safetensors`), `config.json` and
->   `tokenizer/`; the documentation (`README.md` and the other `.md` files, `paper/`, `tasksource_license_audit.csv`); the
+>   `tokenizer/`; the documentation (`README.md` and the other `.md` files, `paper/`, `assets/`, `tasksource_license_audit.csv`); the
 >   training records (`training/*.json`); the model's own evaluation outputs in `eval/`; `SHA256SUMS`.
 > - **MIT** ([LICENSE-CODE](LICENSE-CODE)): the code — `mini_v41/`, `mini_v41_jev/`, `scripts/`, `examples/`,
 >   `training/code/`, `eval/btzsc22/code/` — and the environment and build files: `Dockerfile`, `.dockerignore`, `requirements.txt`,
@@ -54,6 +54,17 @@ full-case (multi-question) protocol.**
 Development codename: mini-v41 / mini-v41-Decisions. The Python packages keep that name (`mini_v41`,
 `mini_v41_jev`), and so do the provenance records in `config.json` and `training/`.
 
+## Architecture
+
+![Ines-1 architecture](assets/architecture.png)
+
+Causal encoder–decoder (CED) mixture-of-experts with Engram n-gram memory: 8 encoder layers with full causal attention,
+then 8 decoder layers that attend, in one softmax, to their own sliding-window K/V (128 tokens) and to a global K/V
+projected once from the encoder output and shared by every decoder layer. All 16 layers use a MoE FFN (16 routed experts,
+top-2, plus 1 shared expert, SwiGLU 1536); Engram is added to the residual before encoder layers 1 and 5. Context is
+2,048 tokens: raw tokens reach every position through the encoder → global K/V path, while features computed inside the
+decoder travel 128 + 7·127 = 1,017 positions. Vector source: [assets/architecture.svg](assets/architecture.svg).
+
 ## Supported decision interface
 
 - **One typed question per model prompt.** The answer is the softmax over the option letters (at most 26) at the
@@ -358,7 +369,7 @@ manifests, data audit, scripts), `eval/` (reports, per-item outputs, speed proto
 `requirements.txt`, `requirements.lock`, `Dockerfile`, [CHANGELOG_CLAIMS.md](CHANGELOG_CLAIMS.md),
 [THIRD_PARTY_DATA_NOTICE.md](THIRD_PARTY_DATA_NOTICE.md), [LICENSING_NOTES.md](LICENSING_NOTES.md),
 [TASKSOURCE_LICENSE_AUDIT.md](TASKSOURCE_LICENSE_AUDIT.md) + `tasksource_license_audit.csv`,
-`paper/` (technical report: PDF + LaTeX sources), [LICENSE](LICENSE) (Apache-2.0), [NOTICE](NOTICE), [LICENSE-CODE](LICENSE-CODE) (MIT), `SHA256SUMS`.
+`assets/` (architecture diagram: PNG + SVG source), `paper/` (technical report: PDF + LaTeX sources), [LICENSE](LICENSE) (Apache-2.0), [NOTICE](NOTICE), [LICENSE-CODE](LICENSE-CODE) (MIT), `SHA256SUMS`.
 
 ## Paper
 

SHA256SUMS

@@ -1,5 +1,5 @@
 97ad8982da911d928c3423f164a38f80608b98289ac4f4b3f292d90ae0ef2262  .dockerignore
-026b23285c5d5e2b117ee5065229bf984d6ae6fad6b21ee1f04c934b7cad1d57  .gitattributes
+f958f33c67320596595c21e8e540481a03456fac85586dbf26f0542d02684f82  .gitattributes
 862263fa1f46c20f0d1e4dac5ffcc75abd55c08211b2c3864c5f8764b9d87793  .gitignore
 4463cc803842f7fe2406ce4e15039ff73842826d0f0c222f6857f73f8da7f54a  CHANGELOG_CLAIMS.md
 e4a1aeafc1f6c01f8fff69aef8a2518631428f3ec96db4a9a56d7bb296cce95f  Dockerfile
@@ -8,9 +8,11 @@ cfc7749b96f63bd31c3c42b5c471bf756814053e847c10f3eb003417bc523d30  LICENSE
 6a7ddcad077839a44697c005ea7e6f455caecaa8054412913108d6e21a1c9dbe  LICENSE-CODE
 121b31536257bd7147d9c9b054f6fb4e2bb68734024415c081ac0b50c7cf48f1  LICENSING_NOTES.md
 1faada60db1ccfc073ae063fc474d9f75cba0c9d2683d03ad66b2074efeb166a  NOTICE
-d74ebe08f28c4f95dbf275fa9ff015c4f6ca7aacf59e3f65add1e15c8572de3c  README.md
+8b3d5abc94f861cd0c3622a42f3403c7bfb33938009a629b05c88a9fad3f2d2a  README.md
 6ec8a2bb55f9331d77b9adbeae7fd71eeb5614a988cd21eee18c7bde775df5ba  TASKSOURCE_LICENSE_AUDIT.md
 1a4df8f10bd30b01e6acc19ef2d935faee5485f0c25452ff64ba6c8598fabd97  THIRD_PARTY_DATA_NOTICE.md
+9708385cf095b543e009b0dd6013aa9985ef68a5b05c5ebef4c14c6a7f97cfa2  assets/architecture.png
+1c641817fa2f3c2a64a2f52f48e65148aa907e4d2e3df68ccd21f6312666fc69  assets/architecture.svg
 757003b29ec1d935f236c06ea1e39cf9362c43de358f8b3df74326f8bfdbd40d  config.json
 8f296d2c1dbe85d285bb45cd80d841e733f84168ff130daf825933917684993b  eval/btzsc22/PROTOCOL.md
 5cc7f2cd08668abe6ac5dd5455e61bc2315e556e26c45906150426810fd49455  eval/btzsc22/REPORT.md