Evaluating LLMs’ Assessment of Mixed-Context Hallucination Through the Lens of Summarization

This paper evaluates large language models’ ability to assess mixed-context hallucination through summarization tasks. We analyze how LLMs perform in detecting and evaluating hallucination when presented with mixed or conflicting information contexts.

The research contributes to understanding LLMs’ self-assessment capabilities and their limitations in detecting their own hallucination patterns.

Recommended citation: S Qi, R Cao, Y He, Z Yuan. (2025). "Evaluating LLMs Assessment of Mixed-Context Hallucination Through the Lens of Summarization." Findings of the Association for Computational Linguistics: ACL 2025.
Download Paper