9 papers
A Multi-Agent Perception-Action Alliance for Efficient Long Video Reasoning
Yichang Xu, Gaowen Liu, Ramana Rao Kompella +6
This paper presents a multi-agent perception-action exploration alliance, dubbed A4VL, for efficient long-video reasoning. A4VL operates in a multi-round perception-action explorat…
Surgery: Mitigating Harmful Fine-Tuning for Large Language Models via Attention Sink
Guozhi Liu, Weiwei Lin, Tiansheng Huang +4
Harmful fine-tuning can invalidate safety alignment of large language models, exposing significant safety risks. In this paper, we utilize the attention sink mechanism to mitigate…
Mitigating Safety Tax via Distribution-Grounded Refinement in Large Reasoning Models
Yingsha Xie, Tiansheng Huang, Enneng Yang +5
Safety alignment incurs safety tax that perturbs a large reasoning model's (LRM) general reasoning ability. Existing datasets used for safety alignment for an LRM are usually const…
R1-Compress: Long Chain-of-Thought Compression via Chunk Compression and Search
Yibo Wang, Haotian Luo, Huanjin Yao +8
Chain-of-Thought (CoT) reasoning enhances large language models (LLMs) by enabling step-by-step problem-solving, yet its extension to Long-CoT introduces substantial computational…
Pharmacist: Safety Alignment Data Curation for Large Language Models against Harmful Fine-tuning
Guozhi Liu, Qi Mu, Tiansheng Huang +4
Harmful fine-tuning issues present significant safety challenges for fine-tuning-as-a-service in large language models. Existing alignment-stage defenses, e.g., Vaccine, Repnoise,…
Gradient Surgery for Safe LLM Fine-Tuning
Biao Yi, Jiahao Li, Baolei Zhang +4
Fine-tuning-as-a-Service introduces a critical vulnerability where a few malicious examples mixed into the user's fine-tuning dataset can compromise the safety alignment of Large L…