Mathematical Reasoning on Math Family Average
67AccuracyDASD
Evaluation Results
| Method | Links | |
|---|---|---|
| DASDBackbone=Qwen3-8B2026.05 | 67 | |
| GRPOBackbone=Qwen3-8B2026.05 | 60.4 | |
| DASDBackbone=Olmo-8B2026.05 | 54.2 | |
| GRPOBackbone=Olmo-8B2026.05 | 49.4 | |
| DASDBackbone=Llama-8B2026.05 | 49.1 | |
| GRPOBackbone=Llama-8B2026.05 | 44.7 | |
| BaseBackbone=Qwen3-8B2026.05 | 38.7 | |
| OPSDBackbone=Qwen3-8B2026.05 | 38.2 | |
| OPSDBackbone=Olmo-8B2026.05 | 34.5 | |
| BaseBackbone=Olmo-8B2026.05 | 33.8 | |
| OPSDBackbone=Llama-8B2026.05 | 31.4 | |
| BaseBackbone=Llama-8B2026.05 | 31.1 |