License : NRSDOCS

1. BASE MODEL ATTRIBUTION & ORIGINAL CREDITS

This model is fundamentally built upon the "Qwen 3.5 9B" architecture and 
pre-trained weights. 

- Original Creator: Alibaba Qwen Team
- Base License: Apache-2.0
- Attribution: All rights to the underlying architecture, tokenization, and 
  pre-trained base weights remain exclusively with the original creators. 
  SKT AI LABS does not claim ownership of the base Qwen 3.5 model.


2. SKT NRS SPECIFIC TUNING & ENHANCEMENT CLAIM

SKT AI LABS has performed extensive, high-quality Supervised Fine-Tuning (SFT) 
and proprietary "Neural Reasoning System" (NRS) boosting on this base model. 
The following enhancements are the exclusive intellectual effort and property 
of SKT AI LABS:

A. Neural Reasoning System (NRS) Boosting:
   - Implementation of proprietary 100x Reasoning Capacity algorithms.
   - Advanced Chain-of-Thought (CoT) structuring using internal NRS tools.
   - "Thinking Mode" optimization for deep logical deduction.

B. Specific Task Tuning (SFT):
   - High-quality SFT on ~ specialized reasoning samples.
   - Enhanced Tool Calling capabilities (Python Executor & Web Search).
   - Optimization for 1 Million Token Context Window (YaRN RoPE Scaling).

C. Performance Optimization:
   - Latency reduction and inference speed boosts for consumer hardware.
   - Coherence and instruction-following refinements specific to SKT standards.


3. OWNERSHIP OF TUNED WEIGHTS & METHODOLOGY

While the base architecture is attributed to its original creators, the 
specific tuned weights, the NRS reasoning configuration, and the specialized 
performance boosts released in this repository are the result of SKT AI LABS' 
dedicated research and tuning efforts. 

We hold the rights to the specific tuning methodology, the NRS data pipelines, 
and the enhanced model configuration. Users utilizing this model acknowledge 
the significant value added by SKT AI LABS' specific tuning.


4. LICENSE TERMS & USAGE RIGHTS

This specific tuned version is released under the Apache-2.0 License. 
Users are permitted to:
- Use the model for commercial and non-commercial purposes.
- Modify and distribute the tuned weights.
- Integrate the model into applications.

CONDITIONS:
- Users must adhere to the base model's original license terms.
- Proper attribution must be given to both the "Alibaba Qwen Team" (for the 
  base) and "SKT AI LABS" (for the NRS Specific Tuning and Boosting).


5. LIMITATION OF LIABILITY

THIS MODEL IS PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND. 
SKT AI LABS IS NOT RESPONSIBLE FOR ANY MISUSE, HALLUCINATIONS, OR OUTPUTS 
GENERATED BY THIS MODEL. THE HIGH-REASONING CAPABILITIES ARE A RESULT OF 
SPECIFIC TUNING AND SHOULD BE USED WITH APPROPRIATE VERIFICATION IN 
CRITICAL APPLICATIONS (MEDICAL, LEGAL, FINANCIAL).

6. DISCLAIMER

This tuned model is released under the NRS DOCS License. Users are free to 
deploy, modify, and distribute this model, provided they:
1. Adhere to the base model's original license terms.
2. Acknowledge SKT AI LABS for the specific NRS tuning and reasoning enhancements.

DISCLAIMER: This model is provided "AS IS." While optimized for high-quality 
reasoning, users should verify critical outputs. SKT AI LABS is not liable for 
misuse or hallucinations.

7. CONTACT & SUPPORT


For inquiries regarding NRS Tuning, Specific SFT requests, or Enterprise 
Boosting services, please contact SKT AI LABS Or Emails Us Here support@sktailabs.in, 
Info@sktailabs.in.

© 2026 SKT AI LABS. All Rights Reserved for NRS Specific Tuning.