Each hallucination detection method in the Cleanlab benchmark takes a user query, retrieved context, and LLM response as input and returns a score between 0 and 1 indicating the likelihood of hallucination.

Authors

• Person: Not available Organization: Cleanlab
  Benchmarking Hallucination Detection Methods in RAG - Cleanlab

Sources

• Benchmarking Hallucination Detection Methods in RAG - Cleanlab cleanlab.ai Cleanlab via serper

Referenced by nodes (1)

• Cleanlab entity