Language Models Cannot Reliably Distinguish Belief from Knowledge and Fact

Details

Author(s):
Publish Date:
2025
Format:
Journal Article
Citation(s):
  • Daniel E. Ho, Mirac Suzgun, Tayfun Gur, Federico Bianchi, Thomas Icard, Dan Jurafsky & James Zou, Language Models Cannot Reliably Distinguish Belief from Knowledge and Fact, Nature: Mach. Intel. (forthcoming 2025).

Other Publications By

Daniel E. Ho