RePPL：通过语义传播和语言生成中的不确定性重新校准困惑度以检测可解释问答幻觉
RePPL: Recalibrating Perplexity by Uncertainty in Semantic Propagation and Language Generation for Explainable QA Hallucination Detection
AI总结：
RePPL通过校准语义传播和语言生成中的不确定性，提升问答幻觉检测性能，生成token级不确定性评分作为解释。
摘要：
大型语言模型（LLMs）已经变得强大，但幻觉仍然是其可信使用的重要障碍。先前的工作通过测量不确定性来提高幻觉检测的能力。但它们无法解释幻觉发生的原因，特别是识别哪些输入部分容易触发幻觉。最近的提示攻击研究表明，在语义传播中存在不确定性，其中注意力机制逐渐将局部token信息融合为高层语义。同时，在语言生成中也存在不确定性，因为其基于概率的选择高层语义用于采样生成。基于此，我们提出了RePPL，通过这两个方面重新校准不确定性测量，将可解释的不确定性评分分配给每个token，并以困惑度式对数平均形式汇总为总评分。实验表明，它在各种QA数据集上实现了最先进的综合检测性能（平均AUC为0.833），并且能够生成token级的不确定性评分作为幻觉的解释。
英文摘要：
Large Language Models (LLMs) have become powerful, but hallucinations remain a vital obstacle to their trustworthy use. Previous works improved the capability of hallucination detection by measuring uncertainty. But they can not explain the provenance behind why hallucinations occur, particularly in identifying which part of the inputs tends to trigger hallucinations. Recent works on the prompt attack indicate that uncertainty exists in semantic propagation, where attention mechanisms gradually fuse local token information into high-level semantics across layers. Meanwhile, uncertainty also emerges in language generation, due to its probability-based selection of high-level semantics for sampled generations. Based on that, we propose RePPL to recalibrate uncertainty measurement by these two aspects, which dispatches explainable uncertainty scores to each token and aggregates in Perplexity-style Log-Average form as a total score. Experiments show that it achieves the best comprehensive detection performance across various QA datasets on advanced models (average AUC of 0.833), and it is capable of producing token-level uncertainty scores as explanations of hallucination.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
