Marks and Tegmark (2023) identified a generalized 'truth direction' within the geometry of Large Language Models, showing that a simple linear probe can consistently distinguish truthful statements across diverse topics and datasets.

Authors

• Person: Not available Organization: arXiv
  A Survey on the Theory and Mechanism of Large Language Models

Sources

• A Survey on the Theory and Mechanism of Large Language Models arxiv.org arXiv via serper

Referenced by nodes (1)

• Large Language Models concept