Mathematics on OlymMATH Easy
15.9Pass@1GSPO RL (vanilla)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GSPO RL (vanilla)temperature=0.6, rollouts=32, Backbone=OLMo3-7B-Instruct-SFT2026.05 | 15.9 | 73 | |
| GSPO RL with novelty bonustemperature=0.6, rollouts=32, Backbone=OLMo3-7B-Instruct-SFT2026.05 | 15.4 | 71 | |
| Puzzle SFT (rejection-sampled DSR traces)temperature=0.6, rollouts=32, Backbone=OLMo3-7B-Instruct-SFT2026.05 | 14.2 | 72 | |
| Base (OLMo3-7B-Instruct-SFT)temperature=0.6, rollouts=322026.05 | 3.2 | 33.9 |