Mathematical Reasoning on AMC 2023 (Accuracy (avg@32), Average)
94.77Accuracy (avg@32)SRaR
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SRaRModel Size=32B2026.05 | 94.77 | — | |
| RaRModel Size=32B2026.05 | 94.22 | — | |
| DS-R1-Distill-Qwen-7B + IC-DAPOBackbone=Qwen-7B, Zero-shot mode=true, Time/Step (s)=315.62026.03 | 93.7 | 67.8 | |
| DAPOModel Size=32B2026.05 | 93.67 | — | |
| DS-R1-Distill-Qwen-7B + CE-GPPOBackbone=Qwen-7B, Zero-shot mode=true, Time/Step (s)=292.52026.03 | 93.3 | 67.3 | |
| SRaRModel Size=8B2026.05 | 92.73 | — | |
| DS-R1-Distill-Qwen-7B + DAPOBackbone=Qwen-7B, Zero-shot mode=true, Time/Step (s)=303.12026.03 | 92.3 | 65.3 | |
| DS-R1-Distill-Qwen-7BBackbone=Qwen-7B, Zero-shot mode=true2026.03 | 90.6 | 61.8 | |
| DAPOModel Size=8B2026.05 | 90.23 | — | |
| DS-R1-Distill-Qwen-7B + GRPOBackbone=Qwen-7B, Zero-shot mode=true, Time/Step (s)=305.62026.03 | 88.8 | 61.4 | |
| RaRModel Size=8B2026.05 | 88.2 | — | |
| RGR-GRPOModel Size=32B2026.05 | 87.42 | — | |
| DS-R1-Distill-Qwen-1.5B + IC-DAPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=477.22026.03 | 86.2 | 56.4 | |
| DS-R1-Distill-Qwen-1.5B + GSPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=437.32026.03 | 85.9 | 55.7 | |
| DS-R1-Distill-Qwen-1.5B + CE-GPPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=464.02026.03 | 85.8 | 55.7 | |
| DS-R1-Distill-Qwen-1.5B + DAPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=459.62026.03 | 84.4 | 53.9 | |
| GRPO-VPSModel Size=8B2026.05 | 83.2 | — | |
| RGR-GRPOModel Size=8B2026.05 | 81.09 | — | |
| DS-R1-Distill-Qwen-1.5B + CISPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=466.32026.03 | 80.9 | 48.8 | |
| GRPOModel Size=32B2026.05 | 80.78 | — | |
| GRPO-VPSModel Size=32B2026.05 | 80.08 | — | |
| DS-R1-Distill-Qwen-1.5B + GRPOBackbone=Qwen-1.5B, Zero-shot mode=true, Time/Step (s)=457.42026.03 | 79.3 | 50.3 | |
| GRPOModel Size=8B2026.05 | 79.06 | — | |
| DS-R1-Distill-Qwen-1.5BBackbone=Qwen-1.5B, Zero-shot mode=true2026.03 | 73.7 | 46.3 | |
| BaselineModel Size=8B2026.05 | 72.03 | — | |
| BaselineModel Size=32B2026.05 | 71.48 | — | |
| GRPOModel=Qwen2.5-Math-7B, G=162026.06 | 68.91 | — | |
| RF++ w/ NTFModel=Qwen2.5-Math-7B, G=12026.06 | 68.91 | — | |
| GRPOModel=Qwen2.5-Math-7B, G=22026.06 | 64.84 | — | |
| C-RF w/ NTFModel=Qwen2.5-Math-7B, G=12026.06 | 61.95 | — | |
| RF++Model=Qwen2.5-Math-7B, G=82026.06 | 58.36 | — | |
| GRPOModel=Qwen2.5-Math-1.5B, G=162026.06 | 57.11 | — | |
| RF++ w/ NTFModel=Qwen2.5-Math-1.5B, G=12026.06 | 55.47 | — | |
| C-RF w/ NTFModel=Qwen2.5-Math-1.5B, G=12026.06 | 55.47 | — | |
| RF++Model=Qwen2.5-Math-1.5B, G=82026.06 | 54.77 | — | |
| GRPOModel=Qwen2.5-Math-1.5B, G=22026.06 | 52.19 | — |