From Argumentation to Labeled Logic Program for LLM Verification

Large language models (LLMs) often generate fluent but incorrect or unsupported statements, commonly referred to as hallucinations. We propose a hallucination detection framework based on a Labeled Logic Program (LLP) architecture that integrates multiple reasoning paradigms, including logic programming, argumentation, probabilistic inference, and abductive explanation. By enriching symbolic rules with semantic, epistemic, and contextual labels and applying discourse-aware weighting, the system prioritizes nucleus claims over peripheral statements during verification. Experiments on three benchmark datasets and a challenging clinical narrative dataset show that LLP consistently outperforms classical symbolic validators, achieving the highest detection accuracy when combined with discourse modeling. A human evaluation further demonstrates that logic-assisted explanations improve both hallucination detection accuracy and user trust. The results suggest that labeled symbolic reasoning with discourse awareness provides a robust and interpretable approach to LLM verification in safety-critical domains.