License: NRSDOCS**

### 1. MODEL CREDITS & ATTRIBUTION

This model is developed by building upon an open-source base model from the Hugging Face ecosystem.

- **Base Model Credits**: All rights, title, and interest in the underlying model architecture, tokenizer, and original pre-trained weights remain exclusively with their respective original creators and organizations.  
- SKT AI LABS expressly does not claim any ownership or authorship over the base model or its foundational components.  
- Full credit and attribution for the base model are hereby acknowledged and reserved for the original developers.

### 2. SKT AI LABS – PROPRIETARY NEURAL REASONING SYSTEM (NRS) TUNING & ENHANCEMENTS

SKT AI LABS has invested significant resources in conducting extensive Supervised Fine-Tuning (SFT) and implementing its proprietary **Neural Reasoning System (NRS)**. These enhancements substantially elevate the model’s reasoning capabilities, coherence, and practical utility.

The following advancements constitute the exclusive intellectual property and original research contributions of **SKT AI LABS**:

**A. Neural Reasoning System (NRS) Framework**
- Proprietary algorithms engineered to deliver up to **100x reasoning capacity** compared to the base model
- Advanced multi-stage Chain-of-Thought (CoT) orchestration supported by internal NRS tooling
- Specialized “Thinking Mode” optimization designed for superior logical deduction, strategic planning, and complex multi-step problem solving
- Enhanced self-consistency and error-correction mechanisms

**B. Specialized Supervised Fine-Tuning (SFT)**
- Curated high-quality training on diverse, domain-specific reasoning datasets
- Advanced tool-calling and function-calling capabilities, including seamless integration with Python Executor, Web Search, and other external tools
- Significant expansion of effective context window (optimized up to **1 Million tokens** through YaRN RoPE Scaling techniques)

**C. Performance, Efficiency & Benchmark Improvements**
- Measurable gains across major industry benchmarks in reasoning, mathematics, coding, general knowledge, and instruction-following tasks
- Optimized inference latency and memory efficiency, enabling superior performance on consumer-grade and edge hardware
- Substantial improvements in response coherence, factual consistency, instruction adherence, and overall output quality

### 3. OWNERSHIP & INTELLECTUAL PROPERTY RIGHTS

While the base model architecture is attributed to its original creators, the **fine-tuned model weights**, **NRS reasoning configuration**, **custom training methodologies**, **data pipelines**, **optimization techniques**, and all resulting performance improvements are the sole and exclusive intellectual property of **SKT AI LABS**.

Users and developers who utilize this model explicitly acknowledge and appreciate the substantial added value provided by SKT AI LABS’ proprietary tuning and enhancements.

### 4. LICENSE TERMS & USAGE RIGHTS

This specifically tuned and enhanced model is released under the **Apache-2.0 License**.

You are granted broad rights to:
- Use the model for both commercial and non-commercial purposes
- Modify, adapt, and create derivative works based on the tuned weights
- Redistribute, share, and integrate the model into any applications, products, services, or research projects

**Required Conditions:**
- Compliance with all applicable terms of the original base model’s license
- Proper attribution to the original base model creators **and** to SKT AI LABS for the NRS-specific tuning, reasoning enhancements, and benchmark improvements

### 5. LIMITATION OF LIABILITY & RESPONSIBLE USE

THIS MODEL IS PROVIDED **“AS IS”** AND **“AS AVAILABLE”** WITHOUT ANY WARRANTY OF ANY KIND, WHETHER EXPRESS, IMPLIED, STATUTORY, OR OTHERWISE.

SKT AI LABS, ITS AFFILIATES, AND CONTRIBUTORS SHALL NOT BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR EXEMPLARY DAMAGES ARISING FROM THE USE, MISUSE, OR INABILITY TO USE THIS MODEL, INCLUDING BUT NOT LIMITED TO HALLUCINATIONS, INACCURATE OUTPUTS, OR ANY RESULTING LOSSES.

**Responsible Use Guidelines:**
- The advanced reasoning capabilities stem from specialized tuning. All outputs — especially in high-stakes domains such as healthcare, legal advice, financial decisions, safety-critical systems, or scientific research — must be independently verified by qualified human experts.
- Users are strongly encouraged to implement appropriate safeguards and human oversight when deploying this model in production environments.

### 6. DISCLAIMER

This model is released under the **NRS DOCS License**. You are free to deploy, modify, and distribute it provided that you:

1. Respect and comply with the original base model’s license terms
2. Provide clear acknowledgment to SKT AI LABS for the proprietary NRS tuning, reasoning system enhancements, and documented benchmark improvements

**General Disclaimer**: Although this model has been optimized for high-performance reasoning and practical utility, it remains an artificial intelligence system subject to inherent limitations. Users assume full responsibility for any outputs generated and for ensuring compliance with all applicable laws and ethical standards.

### 7. CONTACT & SUPPORT

For inquiries related to:
- Custom NRS tuning and domain-specific fine-tuning
- Enterprise-grade enhancements and deployment support
- Collaboration opportunities or licensing discussions

Please contact us at:

**Email**: support@sktailabs.in | info@sktailabs.in

© 2026 SKT AI LABS. All Rights Reserved for NRS Specific Tuning, Enhancements, and Proprietary Technology.