logoalt Hacker News

mattnewtonyesterday at 8:23 PM0 repliesview on HN

This was a big concern for earlier models, but with modern CoT trained models they should be able to come to the conclusion entirely in the thinking trace.