用内置检测器提前发现并修正大模型长文本生成中的幻觉问题
HalluCana: Fixing LLM Hallucination with A Canary Lookahead
- 在模型隐藏层中寻找事实性线索，实现生成过程中的实时幻觉检测
- 生物传记生成质量提升2.5倍，计算开销减少6倍以上
- 适合需要高可信度长文本生成的应用场景
本文提出HalluCana，一种用于检测和纠正大语言模型（LLMs）在长文本生成中事实性幻觉的信标前瞻机制。该方法在生成过程中甚至生成前即能捕捉幻觉迹象，并及时干预。为实现快速检测，我们利用了模型隐藏空间中的内在事实性表征，研究了多种可代理的模型事实性自评指标，并探讨其与预训练阶段上下文熟悉度的关系。在生物传记生成任务中，该方法使生成质量最高提升2.5倍，同时计算消耗减少超过6倍。
原文摘要 · Abstract (English)
In this paper, we present HalluCana, a canary lookahead to detect and correct factuality hallucinations of Large Language Models (LLMs) in long-form generation. HalluCana detects and intervenes as soon as traces of hallucination emerge, during and even before generation. To support timely detection, we exploit the internal factuality representation in the LLM hidden space, where we investigate various proxies to the LLMs' factuality self-assessment, and discuss its relation to the models' context familiarity from their pre-training. On biography generation, our method improves generation quality by up to 2.5x, while consuming over 6 times less compute.
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。