Already Present Beliefs?

by Ari Holtzmanabout 15 hours ago
0

If an LM requires less finetuning data to believe X than Y, is it also more susceptible to believing X in a conversation before finetuning?

Chat

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

@misc{holtzman-already-present-beliefs-2026,
  author = {Holtzman, Ari},
  title = {Already Present Beliefs?},
  year = {2026},
  url = {https://hypogenic.ai/ideahub/idea/POf2Qh3nT3UCo6wq2k8l}
}

Comments (0)

Please sign in to comment on this idea.

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