HALoGEN: Fantastic LLM Hallucinations and Where to Find Them
This addresses the challenge of efficiently evaluating hallucinations in LLMs for researchers and developers, though it is incremental as it builds on existing benchmarking efforts.
The authors tackled the problem of measuring hallucinations in generative large language models (LLMs) by releasing HALoGEN, a benchmark with 10,923 prompts across nine domains and automatic verifiers, finding that even top models have hallucinations up to 86% of generated facts in some domains.
Despite their impressive ability to generate high-quality and fluent text, generative large language models (LLMs) also produce hallucinations: statements that are misaligned with established world knowledge or provided input context. However, measuring hallucination can be challenging, as having humans verify model generations on-the-fly is both expensive and time-consuming. In this work, we release HALoGEN, a comprehensive hallucination benchmark consisting of: (1) 10,923 prompts for generative models spanning nine domains including programming, scientific attribution, and summarization, and (2) automatic high-precision verifiers for each use case that decompose LLM generations into atomic units, and verify each unit against a high-quality knowledge source. We use this framework to evaluate ~150,000 generations from 14 language models, finding that even the best-performing models are riddled with hallucinations (sometimes up to 86% of generated atomic facts depending on the domain). We further define a novel error classification for LLM hallucinations based on whether they likely stem from incorrect recollection of training data (Type A errors), or incorrect knowledge in training data (Type B errors), or are fabrication (Type C errors). We hope our framework provides a foundation to enable the principled study of why generative models hallucinate, and advances the development of trustworthy large language models.