Re:LLMs can't self-correct in reasoning tasks, DeepMind stud...
by Fleusal on Monday October 09, 2023 @03:55PM
(#37824696)
LLMs definitely show signs of understanding, though they might not be the sharpest tools in the shed right now.
[ Parent
| on HN ]
Re:LLMs can't self-correct in reasoning tasks, DeepMind stud...
by nomel on Monday October 09, 2023 @04:01PM
(#37824766)
There’s a very good chance that many of these people have never used something like GPT-4, at all, or outside of treating it like google. This has been the conclusion of every other claim like that, stated with such certainty, that I’ve seen in the wild.