Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2511.03138 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911269623169024 |
|---|---|
| author | Li, Qi Xu, Jianjun Wei, Pingtao Li, Jiu Zhao, Peiqiang Shi, Jiwei Zhang, Xuan Yang, Yanhui Hui, Xiaodong Xu, Peng Shao, Wenqin |
| author_facet | Li, Qi Xu, Jianjun Wei, Pingtao Li, Jiu Zhao, Peiqiang Shi, Jiwei Zhang, Xuan Yang, Yanhui Hui, Xiaodong Xu, Peng Shao, Wenqin |
| contents | With the widespread application of Large Language Models (LLMs), their associated security issues have become increasingly prominent, severely constraining their trustworthy deployment in critical domains. This paper proposes a novel safety response framework designed to systematically safeguard LLMs at both the input and output levels. At the input level, the framework employs a supervised fine-tuning-based safety classification model. Through a fine-grained four-tier taxonomy (Safe, Unsafe, Conditionally Safe, Focused Attention), it performs precise risk identification and differentiated handling of user queries, significantly enhancing risk coverage and business scenario adaptability, and achieving a risk recall rate of 99.3%. At the output level, the framework integrates Retrieval-Augmented Generation (RAG) with a specifically fine-tuned interpretation model, ensuring all responses are grounded in a real-time, trustworthy knowledge base. This approach eliminates information fabrication and enables result traceability. Experimental results demonstrate that our proposed safety control model achieves a significantly higher safety score on public safety evaluation benchmarks compared to the baseline model, TinyR1-Safety-8B. Furthermore, on our proprietary high-risk test set, the framework's components attained a perfect 100% safety score, validating their exceptional protective capabilities in complex risk scenarios. This research provides an effective engineering pathway for building high-security, high-trust LLM applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_03138 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DeepKnown-Guard: A Proprietary Model-Based Safety Response Framework for AI Agents Li, Qi Xu, Jianjun Wei, Pingtao Li, Jiu Zhao, Peiqiang Shi, Jiwei Zhang, Xuan Yang, Yanhui Hui, Xiaodong Xu, Peng Shao, Wenqin Artificial Intelligence With the widespread application of Large Language Models (LLMs), their associated security issues have become increasingly prominent, severely constraining their trustworthy deployment in critical domains. This paper proposes a novel safety response framework designed to systematically safeguard LLMs at both the input and output levels. At the input level, the framework employs a supervised fine-tuning-based safety classification model. Through a fine-grained four-tier taxonomy (Safe, Unsafe, Conditionally Safe, Focused Attention), it performs precise risk identification and differentiated handling of user queries, significantly enhancing risk coverage and business scenario adaptability, and achieving a risk recall rate of 99.3%. At the output level, the framework integrates Retrieval-Augmented Generation (RAG) with a specifically fine-tuned interpretation model, ensuring all responses are grounded in a real-time, trustworthy knowledge base. This approach eliminates information fabrication and enables result traceability. Experimental results demonstrate that our proposed safety control model achieves a significantly higher safety score on public safety evaluation benchmarks compared to the baseline model, TinyR1-Safety-8B. Furthermore, on our proprietary high-risk test set, the framework's components attained a perfect 100% safety score, validating their exceptional protective capabilities in complex risk scenarios. This research provides an effective engineering pathway for building high-security, high-trust LLM applications. |
| title | DeepKnown-Guard: A Proprietary Model-Based Safety Response Framework for AI Agents |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2511.03138 |