Held-out Skill Evaluation on Multi-family n=225 (holdout)
43.5556Solve Rate
GRPO from SFT init
Updated 5d ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.05 | 43.5556 | |
2026.05 | 27.5556 | |
2026.05 | 11.5556 |
GRPO from SFT init
| Method | Links | |
|---|---|---|
2026.05 | 43.5556 | |
2026.05 | 27.5556 | |
2026.05 | 11.5556 |