Search papers, labs, and topics across Lattice.
This paper introduces LaaB, a framework that improves LLM hallucination detection by explicitly modeling the logical consistency between LLM responses and their self-judgments. LaaB uses a "meta-judgment" process to map symbolic self-judgment labels into the feature space, creating a logical bridge where response and meta-judgment labels are either the same or opposite. Experiments across four datasets and four LLMs show LaaB outperforms eight baselines by aligning and integrating dual-view signals via mutual learning.
LLMs' own self-judgments, when logically linked to their response features, can significantly improve hallucination detection.
Large Language Models (LLMs) are prone to factual hallucinations, risking their reliability in real-world applications. Existing hallucination detectors mainly extract micro-level intrinsic patterns for uncertainty quantification or elicit macro-level self-judgments through verbalized prompts. However, these methods address only a single facet of the hallucination, focusing either on implicit neural uncertainty or explicit symbolic reasoning, thereby treating these inherently coupled behaviors in isolation and failing to exploit their interdependence for a holistic view. In this paper, we propose LaaB (Logical Consistency-as-a-Bridge), a framework that bridges neural features and symbolic judgments for hallucination detection. LaaB introduces a"meta-judgment"process to map symbolic labels back into the feature space. By leveraging the inherent logical bridge where response and meta-judgment labels are either the same or opposite based on the self-judgment's semantics, LaaB aligns and integrates dual-view signals via mutual learning and enhances the hallucination detection. Extensive experiments on 4 public datasets, across 4 LLMs, against 8 baselines demonstrate the superiority of LaaB.
