33.
We post-trained a 3B model with RL to beat Opus on spreadsheet retrieval. Faster, cheaper, more accurate.
Editor’s note: imported_from_x_likes
@alexstauffer_ on Backlist
1 appearance on the backlist front page in the last 30 days.
1 appearance on the backlist front page in the last 30 days.
Editor’s note: imported_from_x_likes