News for hackers. Stuff that matters.

A comment on Learning to Reason with LLMs. Parent comment

Re:Learning to Reason with LLMs by vanviegen on Thursday September 12, 2024 @03:49PM (#41524758)
It doesn't do the 'thinking' in English (inference is just math), but it does now verbalize intermediate thoughts in English (or whatever the input language is, presumably), just like humans tend to do.
[ Parent | on HN ]
Re:Learning to Reason with LLMs by konschubert on Friday September 13, 2024 @02:09AM (#41528429)
Agreed. It we never "just autocomplete" unless your definition of "autocomplete" includes "look at the whole body of text".
[ Parent | on HN ]