Hacker News new | ask | show | jobs
by thorum 20 days ago
Interesting that all four models converge on such similar designs, for such short prompts.
1 comments

They were trained pretty much on the same data.