The study 'Survey and analysis of hallucinations in large language models' utilized three primary datasets to analyze hallucination patterns: TruthfulQA, HallucinationEval, and QAFactEval.

Authors

• Person: Not available Organization: Frontiers
  Survey and analysis of hallucinations in large language models

Sources

• Survey and analysis of hallucinations in large language models www.frontiersin.org Frontiers via serper

Referenced by nodes (2)

• TruthfulQA concept
• Survey and analysis of hallucinations in large language models: attribution to prompting strategies or model behavior concept