通过内部状态和结构化推理一致性检测大语言模型中的幻觉
Hallucination Detection via Internal States and Structured Reasoning Consistency in Large Language Models
AI总结：
通过内部状态和结构化推理一致性检测大语言模型中的幻觉，提出统一框架解决检测困境，提升事实和逻辑任务的检测性能。
摘要：
在大型语言模型（LLMs）中检测复杂的幻觉受到一个“检测困境”的阻碍：探测内部状态（内部状态探测）在识别事实性不一致方面表现优异，但在逻辑谬误上却失败；而验证外部化推理（推理链验证）则表现出相反的行为。这种分裂导致了一个任务依赖性的盲区：推理链验证在事实密集型任务如开放式问答中失败，因为推理缺乏基础，而内部状态探测在逻辑密集型任务如数学推理中无效，因为模型自信地错误。我们通过一个统一的框架来弥合这个关键的差距。然而，统一受到两个根本性的挑战的阻碍：信号稀缺障碍，因为粗粒度的符号推理链缺乏可以直接与细粒度内部状态相比较的信号，以及表征对齐障碍，即它们的底层语义空间之间存在深层的不匹配。为了克服这些，我们引入了多路径推理机制以获得更多可比、细粒度的信号，并引入了段意识的时间化交叉注意力模块，以适应性地融合这些现在对齐的表示，定位细微的不一致。在三个多样化的基准和两个领先的LLM上进行了广泛的实验，证明我们的框架在一致性和显著性上都优于强大的基线。我们的代码可用：https://github.com/peach918/HalluDet。
英文摘要：
The detection of sophisticated hallucinations in Large Language Models (LLMs) is hampered by a ``Detection Dilemma'': methods probing internal states (Internal State Probing) excel at identifying factual inconsistencies but fail on logical fallacies, while those verifying externalized reasoning (Chain-of-Thought Verification) show the opposite behavior. This schism creates a task-dependent blind spot: Chain-of-Thought Verification fails on fact-intensive tasks like open-domain QA where reasoning is ungrounded, while Internal State Probing is ineffective on logic-intensive tasks like mathematical reasoning where models are confidently wrong. We resolve this with a unified framework that bridges this critical gap. However, unification is hindered by two fundamental challenges: the Signal Scarcity Barrier, as coarse symbolic reasoning chains lack signals directly comparable to fine-grained internal states, and the Representational Alignment Barrier, a deep-seated mismatch between their underlying semantic spaces. To overcome these, we introduce a multi-path reasoning mechanism to obtain more comparable, fine-grained signals, and a segment-aware temporalized cross-attention module to adaptively fuse these now-aligned representations, pinpointing subtle dissonances. Extensive experiments on three diverse benchmarks and two leading LLMs demonstrate that our framework consistently and significantly outperforms strong baselines. Our code is available: https://github.com/peach918/HalluDet.
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。
