Comment by ScottBurson

1 year ago

> The model genuinely believes it’s giving a correct reasoning chain

The model doesn't "genuinely believe" anything.

0 comments

ScottBurson

Reply

No comments yet

Contribute on Hacker News ↗