Chain of Natural Language Inference for Reducing Large Language Model Hallucinations

Deren Lei, Yaxi Li, Mengya Hu, Mingyu Wang, Xi Yun

Keywords: hallucination detection and reduction, large language models, natural language processing

TL;DR: We propose a generic LLM-based hallucination detection and post-editing mitigation framework targeting ungrounded hallucination in text-to-text generation and achieves high quality results.

Abstract: Large language models (LLMs) can generate fluent natural language texts when given relevant documents as background context. This ability has attracted considerable interest in developing industry applications of LLMs. However, LLMs are prone to generate hallucinations that are not supported by the provided sources. In this paper, we propose a hierarchical framework to detect and mitigate such ungrounded hallucination. Our framework uses Chain of Natural Language Inference (CoNLI) for hallucination detection and hallucination reduction via post-editing. Our approach achieves state-of-the-art performance on hallucination detection and enhances text quality through rewrite, using LLMs without any fine-tuning or domain-specific prompt engineering. We show that this simple plug-and-play framework can serve as an effective choice for hallucination detection and reduction, achieving competitive performance across various contexts.

Submission Number: 88