logoalt Hacker News

embedding-shapeyesterday at 8:41 AM0 repliesview on HN

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