logoalt Hacker News

jrflowersyesterday at 1:23 AM1 replyview on HN

Why would you do that


Replies

embedding-shapeyesterday at 8:41 AM

I thought that was clear, try to show biases in LLMs with a concrete example.