Hacker News new | ask | show | jobs
by rawrawrawrr 648 days ago
Nice! It would be a better benchmark to compare this prompt (w/ gpt-4o, claude) with whatever the original model was compared to.
1 comments
windexh8er 648 days ago
It appears that they have. [0]

[0] https://x.com/mattshumer_/status/1831767017507954808

link