Mathematical Reasoning on GSM-Plus 2400
40.25Pass@1DeepSeek-7B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DeepSeek-7BGroup Size (G)=42026.06 | 40.25 | 60.5 | |
| DeepSeek-R1-Qwen-1.5BGroup Size (G)=162026.06 | 36.52 | 57.86 | |
| DeepSeek-R1-Qwen-1.5BGroup Size (G)=42026.06 | 34.8 | 56.44 | |
| DeepSeek-R1-Qwen-1.5BGroup Size (G)=Base2026.06 | 33.48 | 55.54 | |
| DeepSeek-7BGroup Size (G)=162026.06 | 30.83 | 52.19 | |
| DeepSeek-7BGroup Size (G)=Base2026.06 | 25.64 | 49.25 | |
| Nemotron-4BGroup Size (G)=162026.06 | 20.32 | 40.86 | |
| Nemotron-4BGroup Size (G)=42026.06 | 19.46 | 42.02 | |
| Nemotron-4BGroup Size (G)=Base2026.06 | 9.11 | 21.92 |