一作ACL 2026Poster· Natural Language Processing
S afe C onstellations: Mitigating Over-Refusals in LLM s Through Task-Aware Representation Steering
Utsav Maskey · Sumit Yadav · Mark Dras · Usman Naseem
摘要
LLMs increasingly exhibit over-refusal behavior, where safety mechanisms cause models to reject benign instructions that seemingly resemble harmful content. This phenomenon diminishes utility in production applications that repeatedly rely on common prompt templates or applications that frequentl...