The authors of the survey and analysis of hallucinations in large language models claim this is the first application of Bayesian hierarchical modeling to LLM hallucination analysis, utilizing Markov Chain Monte Carlo (MCMC) sampling to yield credible intervals and posterior distributions.

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)

• Large Language Models concept