analiza aiqbench
This research reframes hallucination detection in large language models as an out-of-distribution problem, enabling training-free, single-sample detectors that significantly improve accuracy on reasoning tasks. By leveraging geometric perspectives from computer vision, the approach offers a scalable pathway to enhance model safety beyond standard question-answering benchmarks.
From Out-of-Distribution Detection to Hallucination Detection: A Geometric View
arXiv:2602.07253v3 Announce Type: replace-cross Abstract: Detecting hallucinations in large language models is a critical open problem with significant implications for safety and reliability. While existing hallucination detection methods achieve strong performance in question-answering tasks, they remain less effective on tasks requiring reasoning. In this work, we revisit hallucination detection through the lens of out-of-distribution (OOD) detection, a well-studied problem in areas like computer vision. Treating next-token prediction in language models as a classification task allows us to apply OOD techniques, provided appropriate modifications are made to account for the structural differences in large language models. We show that OOD-based approaches yield training-free, single-sample-based detectors, achieving strong accuracy in hallucination detection for reasoning tasks. Overall, our work suggests that reframing hallucination detection as OOD detection provides a promising and scalable pathway toward language model safety.
Abstract:Detecting hallucinations in large language models is a critical open problem with significant implications for safety and reliability. While existing hallucination detection methods achieve strong performance in question-answering tasks, they remain less effective on tasks requiring reasoning. In this work, we revisit hallucination detection through the lens of out-of-distribution (OOD) detection, a well-studied problem in areas like computer vision. Treating next-token prediction in language models as a classification task allows us to apply OOD techniques, provided appropriate modifications are made to account for the structural differences in large language models. We show that OOD-based approaches yield training-free, single-sample-based detectors, achieving strong accuracy in hallucination detection for reasoning tasks. Overall, our work suggests that reframing hallucination detection as OOD detection provides a promising and scalable pathway toward language model safety.
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2602.07253 [cs.AI]
(or arXiv:2602.07253v3 [cs.AI] for this version)
Text integral sub licența deschisă a publicației · 144 cuvinte