Mathematical Reasoning on ProcessBench GSM8K 1.0 (test)
96AccuracyQwen2.5-Math-PRM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen2.5-Math-PRMModel Size=72B2025.03 | 96 | 96.1 | 93.8 | 98.5 | |
| Greedy EDU PRMModel Size=72B2025.03 | 94.2 | 95 | 90.9 | 99.5 | |
| Omega PRMModel Size=72B2025.03 | 90.5 | 90.8 | 88.2 | 93.5 | |
| Qwen2.5-Math-PRMModel Size=7B2025.03 | 88.8 | 89.5 | 83.8 | 96 | |
| Math-Shepherd PRMModel Size=72B2025.03 | 74.5 | 80.3 | 67.1 | 100 | |
| Sample EDU PRMModel Size=72B2025.03 | 71 | 77.8 | 63.7 | 100 | |
| Omega PRMModel Size=7B2025.03 | 57.5 | 31 | 84.4 | 19 | |
| Math-Shepherd PRMModel Size=7B2025.03 | 57.2 | 68.2 | 54.5 | 91 | |
| Greedy EDU PRMModel Size=7B2025.03 | 55.2 | 21.8 | 86.2 | 12.5 | |
| Sample EDU PRMModel Size=7B2025.03 | 52.5 | 67.7 | 51.3 | 99.5 |