DelucionQA: Detecting Hallucinations in Domain-specific Question Answering
This addresses the critical issue of hallucination in LLM-generated text for high-reliability applications like customer-facing assistants, but it is incremental as it builds on existing retrieval-augmented methods.
The authors tackled the problem of detecting hallucinations in domain-specific question answering by introducing a new dataset, DelucionQA, and proposing baseline detection methods, though no concrete performance numbers are provided.
Hallucination is a well-known phenomenon in text generated by large language models (LLMs). The existence of hallucinatory responses is found in almost all application scenarios e.g., summarization, question-answering (QA) etc. For applications requiring high reliability (e.g., customer-facing assistants), the potential existence of hallucination in LLM-generated text is a critical problem. The amount of hallucination can be reduced by leveraging information retrieval to provide relevant background information to the LLM. However, LLMs can still generate hallucinatory content for various reasons (e.g., prioritizing its parametric knowledge over the context, failure to capture the relevant information from the context, etc.). Detecting hallucinations through automated methods is thus paramount. To facilitate research in this direction, we introduce a sophisticated dataset, DelucionQA, that captures hallucinations made by retrieval-augmented LLMs for a domain-specific QA task. Furthermore, we propose a set of hallucination detection methods to serve as baselines for future works from the research community. Analysis and case study are also provided to share valuable insights on hallucination phenomena in the target scenario.