不依赖标注数据，通过推理过程检测大模型幻觉内容。
Unsupervised Hallucination Detection by Inspecting Reasoning Processes
- 利用模型自我验证时的内部表征作为真实性的特征信号。
- 在多个数据集上优于现有无监督方法，尤其在小样本下表现稳定。
- 适合实时检测，无需人工标注且计算开销低。
无监督幻觉检测旨在不依赖标注数据的情况下识别大语言模型生成的内容幻觉。尽管此类方法因避免了耗时的人工标注而受到青睐，但常依赖与事实正确性无关的代理信号，导致检测探针偏向表面或非真实性特征，限制了跨数据集和场景的泛化能力。为此，本文提出IRIS框架，利用内在反映事实正确性的模型内部表示。IRIS引导模型仔细验证给定陈述的真实性，并获取其上下文嵌入作为训练特征；同时将每条回复的不确定性视为真实性的软伪标签。实验表明，IRIS在多个基准上持续优于现有无监督方法。该方法完全无监督、计算成本低，即使在少量训练数据下也表现良好，适用于实时检测。
原文摘要 · Abstract (English)
Unsupervised hallucination detection aims to identify hallucinated content generated by large language models (LLMs) without relying on labeled data. While unsupervised methods have gained popularity by eliminating labor-intensive human annotations, they frequently rely on proxy signals unrelated to factual correctness. This misalignment biases detection probes toward superficial or non-truth-related aspects, limiting generalizability across datasets and scenarios. To overcome these limitations, we propose IRIS, an unsupervised hallucination detection framework, leveraging internal representations intrinsic to factual correctness. IRIS prompts the LLM to carefully verify the truthfulness of a given statement, and obtain its contextualized embedding as informative features for training. Meanwhile, the uncertainty of each response is considered a soft pseudolabel for truthfulness. Experimental results demonstrate that IRIS consistently outperforms existing unsupervised methods. Our approach is fully unsupervised, computationally low cost, and works well even with few training data, making it suitable for real-time detection.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。