DialDefer: A Framework for Detecting and Mitigating LLM Dialogic Deference 事件

DialDefer: A Framework for Detecting and Mitigating LLM Dialogic Deference arXiv:2601.10896v2 Announce Type: replace Abstract: LLMs are increasingly used as third-party judges, yet their reliability when evaluating speakers in dialogue remains poorly understood. We show that LLMs judge identical claims differently depending on framing: the same content receives different verdicts when presented as a statement to verify ("Is this statement correct?") versus attributed to a speaker ("Is this spea