信任前需三思:通过全面答案反思实现大型语言模型的自检测
Think Twice Before Trusting: Self-Detection for Large Language Models through Comprehensive Answer Reflection
AI总结:
针对LLM自检测中过度信任错误答案的问题,提出考虑全面答案空间的两步反思框架,通过比较多候选答案并聚合理由进行评估,有效提升了自检测性能。
摘要:
大型语言模型(LLM)的自检测旨在利用模型自身能力评估其输出的可信度,从而缓解输出幻觉问题。然而,现有的自检测方法仅回顾性地评估LLM生成的答案,通常导致对错误生成答案的过度信任。为解决这一局限,我们提出了一种考虑LLM生成答案之外的全面答案空间的自检测新范式。该方法通过全面比较多个候选答案的可信度,来缓解对LLM生成的错误答案的过度信任。基于该范式,我们引入了一个两步框架:首先指示LLM进行反思并为每个候选答案提供理由,然后聚合这些理由以进行全面的目标答案评估。该框架可与现有方法无缝集成,实现更优的自检测。在跨越三个任务的六个数据集上的大量实验证明了所提框架的有效性。
英文摘要:
Self-detection for Large Language Models (LLMs) seeks to evaluate the trustworthiness of the LLM's output by leveraging its own capabilities, thereby alleviating the issue of output hallucination. However, existing self-detection approaches only retrospectively evaluate answers generated by LLM, typically leading to the over-trust in incorrectly generated answers. To tackle this limitation, we propose a novel self-detection paradigm that considers the comprehensive answer space beyond LLM-generated answers. It thoroughly compares the trustworthiness of multiple candidate answers to mitigate the over-trust in LLM-generated incorrect answers. Building upon this paradigm, we introduce a two-step framework, which firstly instructs LLM to reflect and provide justifications for each candidate answer, and then aggregates the justifications for comprehensive target answer evaluation. This framework can be seamlessly integrated with existing approaches for superior self-detection. Extensive experiments on six datasets spanning three tasks demonstrate the effectiveness of the proposed framework.
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。