A New Benchmark and Reverse Validation Method for Passage-level Hallucination Detection
This addresses the critical issue of LLM hallucinations in mission-critical applications, though it appears incremental by focusing on passage-level detection rather than sentence-level.
The authors tackled the problem of detecting factual errors (hallucinations) in Large Language Models by proposing a self-check reverse validation method and creating a passage-level benchmark called PHD. Their method outperformed existing zero-resource baselines on two datasets while using fewer tokens and less time.
Large Language Models (LLMs) have shown their ability to collaborate effectively with humans in real-world scenarios. However, LLMs are apt to generate hallucinations, i.e., makeup incorrect text and unverified information, which can cause significant damage when deployed for mission-critical tasks. In this paper, we propose a self-check approach based on reverse validation to detect factual errors automatically in a zero-resource fashion. To facilitate future studies and assess different methods, we construct a hallucination detection benchmark named PHD, which is generated by ChatGPT and annotated by human annotators. Contrasting previous studies of zero-resource hallucination detection, our method and benchmark concentrate on passage-level detection instead of sentence-level. We empirically evaluate our method and existing zero-resource detection methods on two datasets. The experimental results demonstrate that the proposed method considerably outperforms the baselines while costing fewer tokens and less time. Furthermore, we manually analyze some hallucination cases that LLM failed to capture, revealing the shared limitation of zero-resource methods.