Which is why sometimes it takes three “Nope, that’s wrong” responses to get proper output.

My theory is that text-based LLMs are intuitive / one-pass thinkers and if a problem requires iteration you’ll need a few passes.
LLMs act like a single brain region but are part of getting to AGI
129
7
Paul Pallaghy, PhD
Anthony (Tony/Pcunix) Lawrence 👀
·Follow
May 24, 2024
--
Which is why sometimes it takes three “Nope, that’s wrong” responses to get proper output.
--
--
Written by Anthony (Tony/Pcunix) Lawrence 👀4.6K Followers
·142 Following
Retired Unix Consultant. I write tech and humor mostly but sometimes other things. See my Lists if your interests are specific.
Responses (1)
Help
Status
About
Careers
Press
Blog
Privacy
Terms
Text to speech
Teams