A Lie Detector Test for Language Models: Reading Knowledge a Model Won't Reveal· 语言模型的说谎检测
Large language models can hold knowledge they do not report. A model may sandbag on a capability evaluation, or answer against what it internally knows, and its outputs alone cannot tell whether it is hiding an answer or simply does not have one. We borrow the Concealed Information Test, a forensic method that identifies guilty knowledge by presenting a suspect with the true detail among plausible decoys and measuring a stronger response to the item they recognize. Our method, Probe of Internal Recognition (PIR), does the same inside a model. It presents a question with its candidate answers and reads, from the model's internal states, which candidate the model recognizes as correct. PIR is reference-free, needing no honest reference model and no labeled truth corpus. Across eight models f
语言模型的内部知识识别方法，可检测模型是否隐瞒答案。
- 核心方法
- 借鉴隐蔽信息测试，提出内部识别探测（PIR），通过向模型提供问题及其候选答案来识别模型内部认为正确答案的反应强度。
- 适合谁读
- 研究者
- 要解决的问题
- 大型语言模型可能隐瞒其内部已知的信息，仅通过输出无法判断模型是否持有但未透露相关信息。
- 关键实验
- 未提供
- 主要贡献
- PIR 方法无需参考模型或标注的真实语料库，可在模型内部直接识别其认识的答案。
- 意义与局限
- 为评估语言模型的真实知识水平提供新工具，但可能受限于不同模型的内部机制差异。