一作ACL 2026Poster· Natural Language Processing
Please refuse to answer me! Mitigating Over-Refusal in Large Language Models via Adaptive Contrastive Decoding
Yupeng Qi · Ziyu Lyu · Lixin Cui · Lu Bai · Feng Xia
摘要
Safety-aligned large language models (LLMs) often generate refusal responses to harmless queries due to the over-refusal problem. However, existing methods for mitigating over-refusal cannot maintain a low refusal ratio for harmless queries while keeping a high refusal ratio for malicious ones. I...