The paper 'Internal Consistency and Self-Feedback in Large Language Models: A Survey' proposes an 'Internal Consistency' framework to enhance reasoning and alleviate hallucinations, which consists of three components: Self-Evaluation, Internal Consistency Signal, and Self-Update.

Authors

• Person: Not available Organization: GitHub
  EdinburghNLP/awesome-hallucination-detection - GitHub

Sources

• EdinburghNLP/awesome-hallucination-detection - GitHub github.com GitHub via serper

Referenced by nodes (2)

• Large Language Models concept
• LLM-as-a-judge concept