Data analysis step verification on DABStep
77.78Easy ScoreDataPRM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DataPRMParams=4B, N=162026.04 | 77.78 | 33.86 | 40.89 | |
| Majority VoteParams=-, N=82026.04 | 76.39 | 29.63 | 37.11 | |
| Majority VoteParams=-, N=162026.04 | 76.39 | 30.69 | 38 | |
| LLM-as-a-judgeParams=-, N=82026.04 | 76.39 | 27.51 | 35.33 | |
| Self-RewardingParams=-, N=82026.04 | 76.39 | 30.16 | 37.56 | |
| Self-RewardingParams=-, N=162026.04 | 76.39 | 32.8 | 39.77 | |
| Qwen2.5-Math-PRM-72BParams=72B, N=82026.04 | 76.39 | 22.75 | 31.33 | |
| DataPRMParams=4B, N=82026.04 | 76.39 | 32.8 | 39.77 | |
| Majority VoteParams=-, N=42026.04 | 75 | 26.98 | 34.66 | |
| LLM-as-a-judgeParams=-, N=42026.04 | 75 | 25.13 | 33.11 | |
| LLM-as-a-judgeParams=-, N=162026.04 | 75 | 29.63 | 36.89 | |
| Self-RewardingParams=-, N=42026.04 | 75 | 28.04 | 35.55 | |
| Math-Shepherd-PRM-7BParams=7B, N=42026.04 | 75 | 23.28 | 31.56 | |
| Math-Shepherd-PRM-7BParams=7B, N=82026.04 | 75 | 23.28 | 31.56 | |
| Math-Shepherd-PRM-7BParams=7B, N=162026.04 | 75 | 19.31 | 28.22 | |
| Qwen2.5-Math-PRM-7BParams=7B, N=42026.04 | 75 | 20.9 | 29.56 | |
| ReasonFlux-PRM-7BParams=7B, N=162026.04 | 75 | 13.76 | 23.56 | |
| ThinkPRMParams=14B, N=42026.04 | 75 | 25.13 | 33.11 | |
| ThinkPRMParams=14B, N=82026.04 | 75 | 24.34 | 32.45 | |
| GenPRMParams=32B, N=42026.04 | 75 | 24.6 | 32.66 | |
| Qwen2.5-Math-PRM-72BParams=72B, N=162026.04 | 75 | 20.37 | 29.11 | |
| DataPRMParams=4B, N=42026.04 | 75 | 29.89 | 37.11 | |
| Qwen2.5-Math-PRM-7BParams=7B, N=162026.04 | 73.61 | 14.55 | 24 | |
| ReasonFlux-PRM-7BParams=7B, N=42026.04 | 73.61 | 20.63 | 29.11 | |
| GenPRMParams=32B, N=82026.04 | 73.61 | 25.4 | 33.11 | |
| GenPRMParams=32B, N=162026.04 | 73.61 | 26.72 | 34.22 | |
| Qwen2.5-Math-PRM-72BParams=72B, N=42026.04 | 73.61 | 21.96 | 30.22 | |
| Qwen2.5-Math-PRM-7BParams=7B, N=82026.04 | 72.22 | 18.25 | 26.89 | |
| ReasonFlux-PRM-7BParams=7B, N=82026.04 | 72.22 | 17.99 | 26.67 | |
| ThinkPRMParams=14B, N=162026.04 | 72.22 | 26.72 | 34 |