yup: 5 real bugs
Replays of yup: 5 real bugs: every model on every bug, side by side, with the result of each run.
Result
- Claude Haiku 5.5 fixed 3 of 5 bugs and scored null of null points (6 min 38 s).
- GPT-6 Luna fixed 2 of 5 bugs and scored null of null points (5 min 49 s).
Published 2026-11-11.