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 the_gipsy on Monday October 09, 2023 @04:37PM (#37825194)
Do "relationships between concepts" have any meaning, for us humans, if we can't express them?
[ Parent | on HN ]
Re:LLMs can't self-correct in reasoning tasks, DeepMind stud... by cj on Monday October 09, 2023 @04:40PM (#37825229)
Humans can express them.

But we rarely do, outside of (maybe) therapy.

[ Parent | on HN ]