News for hackers. Stuff that matters.

A comment on LLMs can't self-correct in reasoning tasks, DeepMind study finds. Parent comment

Re:LLMs can't self-correct in reasoning tasks, DeepMind stud... by fragmede on Tuesday October 10, 2023 @10:05AM (#37832265)
> Please don't post comments saying that HN is turning into Reddit. It's a semi-noob illusion, as old as the hills.

https://news.ycombinator.com/newsguidelines.html

[ Parent | on HN ]
Re:LLMs can't self-correct in reasoning tasks, DeepMind stud... by mvdtnz on Tuesday October 10, 2023 @03:12PM (#37836168)
I didn't say HN is turning into Reddit, I criticised one specific comment. Thanks for your input but you can take it elsewhere.
[ Parent | on HN ]