Are LLMs being trained using LessWrong text? — LessWrong

I wonder if there’s a clear evidence that LessWrong text has been included in LLM training.

Claude seems generally aware of LessWrong, but it’s difficult to distinguish between “this model has been trained on text that mentions LessWrong” and “this model has been trained on text from LessWrong”

Related discussion here, about preventing inclusion:

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top