Are LLMs better at predicting themselves than people are?

0

If you ask an LLM about aspects of how it will do something (e.g. style, mentions, or statements in an essay) is it better at predicting itself than humans? I have no idea how to test this, since humans will anchor on their responses.

llms self-prediction

Chat

If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:

@misc{holtzman-are-llms-better-2026,
  author = {Holtzman, Ari},
  title = {Are LLMs better at predicting themselves than people are?},
  year = {2026},
  url = {https://hypogenic.ai/ideahub/idea/LwQt5D8FCknNYUklsUEO}
}

Comments (0)

Please sign in to comment on this idea.

No comments yet. Be the first to share your thoughts!