Comment by andai 2 days ago Interesting. Wasn't their previous attempt (Phi) trained mostly on synthetic data? 0 comments andai Reply No comments yet Contribute on Hacker News ↗
No comments yet
Contribute on Hacker News ↗