Math Reasoning on AMC23 (mean@12)
92.7Mean Score @12RLCSD
Evaluation Results
| Method | Links | |
|---|---|---|
| RLCSDModel=Olmo-3-7B2026.06 | 92.7 | |
| RLSDModel=Olmo-3-7B2026.06 | 92.6 | |
| GRPOModel=Olmo-3-7B2026.06 | 92.4 | |
| OPSDModel=Olmo-3-7B2026.06 | 92.2 | |
| SRPOModel=Olmo-3-7B2026.06 | 92.1 | |
| SDPOModel=Olmo-3-7B2026.06 | 91.6 | |
| BaseModel=Olmo-3-7B2026.06 | 91.2 | |
| RLCSDModel=Qwen3-8B2026.06 | 90.8 | |
| OPSDModel=Qwen3-8B2026.06 | 90.4 | |
| RLCSDModel=Qwen3-4B2026.06 | 90.1 | |
| GRPOModel=Qwen3-8B2026.06 | 90.1 | |
| OPSDModel=Qwen3-4B2026.06 | 89.4 | |
| GRPOModel=Qwen3-4B2026.06 | 89.1 | |
| BaseModel=Qwen3-8B2026.06 | 88.8 | |
| SDPOModel=Qwen3-8B2026.06 | 88.8 | |
| RLSDModel=Qwen3-8B2026.06 | 88.7 | |
| BaseModel=Qwen3-4B2026.06 | 88.6 | |
| SDPOModel=Qwen3-4B2026.06 | 88.3 | |
| SRPOModel=Qwen3-8B2026.06 | 88.3 | |
| SRPOModel=Qwen3-4B2026.06 | 87.9 | |
| RLSDModel=Qwen3-4B2026.06 | 86.9 | |
| RLCSDModel=Qwen3-1.7B2026.06 | 77.2 | |
| GRPOModel=Qwen3-1.7B2026.06 | 76.6 | |
| OPSDModel=Qwen3-1.7B2026.06 | 76.3 | |
| BaseModel=Qwen3-1.7B2026.06 | 74.1 | |
| RLSDModel=Qwen3-1.7B2026.06 | 73.9 | |
| SRPOModel=Qwen3-1.7B2026.06 | 73.2 | |
| SDPOModel=Qwen3-1.7B2026.06 | 72.9 |