Xi Wu
2 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper challenges the assumption that LLM safety is a binary threshold, proposing that safety failures occur in an 'instability region' and introducing Furina, a transferable attack that exploits this region.
The paper demonstrates that audio-language models often ignore conflicting audio evidence in favor of text, and proposes a training-free decoding rule, GACL, that significantly improves faithfulness by correcting this arbitration bias.
Papers
Beyond Text Following: Repairable Arbitration Reversals in Audio-Language Models
Yichen Gao, Yiqun Zhang, Zijing Wang, Yujia Li +6 more
The paper demonstrates that audio-language models often ignore conflicting audio evidence in favor of text, and proposes a training-free decoding rule, GACL, that significantly improves faithfulness b…