SINdex：用于大语言模型幻觉检测的语义不一致性指标
SINdex: Semantic INconsistency Index for Hallucination Detection in LLMs
AI总结：
本文提出基于语义聚类的幻觉检测框架SINdex，利用句子嵌入和层次聚类提升LLMs幻觉检测准确性，在多个问答数据集上AUROC最高提升9.3%。
摘要：
大语言模型（LLMs）越来越多地部署于各个领域，但它们容易生成事实错误的输出——通常称为“幻觉”。在现有的缓解策略中，基于不确定性的方法因其易于实现、不依赖外部数据且与标准LLMs兼容而特别具有吸引力。在这项工作中，我们引入了一种新颖且可扩展的基于不确定性的语义聚类框架，用于自动化幻觉检测。我们的方法利用句子嵌入和层次聚类，并结合新提出的不一致性度量SINdex，以产生更同质的聚类，并在各种LLMs上更准确地检测幻觉现象。在知名的开放书和闭书问答数据集上的评估表明，与最先进技术相比，我们的方法在AUROC上实现了高达9.3%的提升。广泛的消融研究进一步验证了我们框架中每个组件的有效性。
英文摘要：
Large language models (LLMs) are increasingly deployed across diverse domains, yet they are prone to generating factually incorrect outputs - commonly known as "hallucinations." Among existing mitigation strategies, uncertainty-based methods are particularly attractive due to their ease of implementation, independence from external data, and compatibility with standard LLMs. In this work, we introduce a novel and scalable uncertainty-based semantic clustering framework for automated hallucination detection. Our approach leverages sentence embeddings and hierarchical clustering alongside a newly proposed inconsistency measure, SINdex, to yield more homogeneous clusters and more accurate detection of hallucination phenomena across various LLMs. Evaluations on prominent open- and closed-book QA datasets demonstrate that our method achieves AUROC improvements of up to 9.3% over state-of-the-art techniques. Extensive ablation studies further validate the effectiveness of each component in our 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。
