The attribution framework categorizes hallucinations in Large Language Models into four types: prompt-dominant, model-dominant, mixed-origin, or unclassified.

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)

• Large Language Models concept
• hallucination concept