Model-based mitigation techniques require more infrastructure and training resources than prompt engineering but offer more robust mitigation for model-intrinsic hallucinations.

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)

• prompt engineering concept