A Novel Approach to Eliminating Hallucinations in Large Language Model-Assisted Causal Discovery
This addresses the issue of unreliable model selection for researchers and practitioners using LLMs as substitutes for human experts in causal discovery, though it is incremental as it builds on existing RAG and debate techniques.
The paper tackles the problem of hallucinations in large language models (LLMs) used for causal discovery, showing that hallucinations exist and proposing two methods: Retrieval Augmented Generation (RAG) and a novel debate-based approach with multiple LLMs and an arbiter, both achieving comparable reductions in hallucinations.
The increasing use of large language models (LLMs) in causal discovery as a substitute for human domain experts highlights the need for optimal model selection. This paper presents the first hallucination survey of popular LLMs for causal discovery. We show that hallucinations exist when using LLMs in causal discovery so the choice of LLM is important. We propose using Retrieval Augmented Generation (RAG) to reduce hallucinations when quality data is available. Additionally, we introduce a novel method employing multiple LLMs with an arbiter in a debate to audit edges in causal graphs, achieving a comparable reduction in hallucinations to RAG.