Several established hallucination detection methods show performance drops of up to 45.9% when assessed using human-aligned metrics like LLM-as-Judge compared to traditional metrics.

Authors

• Person: Not available Organization: arXiv
  The Illusion of Progress: Re-evaluating Hallucination Detection in ...

Sources

• The Illusion of Progress: Re-evaluating Hallucination Detection in ... arxiv.org arXiv via serper
• Re-evaluating Hallucination Detection in LLMs - arXiv arxiv.org arXiv via serper

Referenced by nodes (2)

• hallucination detection concept
• LLM-as-a-judge concept