Using auxiliary classifiers or LLMs-as-judges to score and post-edit generated content is a promising direction for hallucination mitigation, as identified by Liu et al. (2023).

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 (1)

• LLM-as-a-judge concept