The strength with which a fact is encoded in a large language model is mediated by complex data pipeline choices, making it difficult to predict or audit the relationship between the frequency of a fact in raw web data and its encoding in the model.

Authors

• Person: M. Brenndoerfer Organization: mbrenndoerfer.com
  Hallucination Causes: Why Language Models Fabricate Facts

Sources

• Hallucination Causes: Why Language Models Fabricate Facts mbrenndoerfer.com M. Brenndoerfer · mbrenndoerfer.com via serper

Referenced by nodes (1)

• Large Language Models concept