Math Reasoning on Math Reasoning Long Q, Long A (test)
0.65Pass@1GEMMA2-9B-IT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GEMMA2-9B-ITWatermark=None2026.02 | 0.65 | — | — | |
| WorldCupBackbone=GEMMA2-9B-IT, k=12026.02 | 0.63 | 0.676 | 0.659 | |
| MPACBackbone=GEMMA2-9B-IT2026.02 | 0.625 | 0.67 | 0.67 | |
| WorldCupBackbone=GEMMA2-9B-IT, k=22026.02 | 0.6 | 0.678 | 0.755 | |
| BiMarkBackbone=GEMMA2-9B-IT2026.02 | 0.59 | 0.676 | 0.706 | |
| LLAMA3.1-8B-ITWatermark=None2026.02 | 0.5 | — | — | |
| WorldCupBackbone=LLAMA3.1-8B-IT, k=12026.02 | 0.468 | 0.814 | 0.917 | |
| WorldCupBackbone=LLAMA3.1-8B-IT, k=22026.02 | 0.445 | 0.892 | 0.947 | |
| BiMarkBackbone=LLAMA3.1-8B-IT2026.02 | 0.42 | 0.868 | 0.922 | |
| WorldCupBackbone=MINISTRAL-8B-IT, k=12026.02 | 0.355 | 0.777 | 0.854 | |
| MPACBackbone=LLAMA3.1-8B-IT2026.02 | 0.343 | 0.856 | 0.927 | |
| MINISTRAL-8B-ITWatermark=None2026.02 | 0.34 | — | — | |
| WorldCupBackbone=MINISTRAL-8B-IT, k=22026.02 | 0.315 | 0.898 | 0.953 | |
| BiMarkBackbone=MINISTRAL-8B-IT2026.02 | 0.31 | 0.767 | 0.877 | |
| MPACBackbone=MINISTRAL-8B-IT2026.02 | 0.265 | 0.785 | 0.879 |