Mathematical Reasoning on AIME 24 (Pass@1, Token count)
80Pass@1RED
Evaluation results
| Method | Links | ||
|---|---|---|---|
| REDModel=Qwen3-32B-thinking2026.04 | 80 | 5,793 | |
| AlphaOneModel=Qwen3-32B-thinking2026.04 | 78.9 | 8,007 | |
| GRPOModel=Qwen3-32B-thinking2026.04 | 78.9 | 11,934 | |
| VanillaModel=Qwen3-32B-thinking2026.04 | 77.8 | 10,677 | |
| DASTModel=Qwen3-32B-thinking2026.04 | 77.8 | 5,981 | |
| S-GRPOModel=Qwen3-32B-thinking2026.04 | 77.8 | 6,040 | |
| Think or NotModel=Qwen3-32B-thinking2026.04 | 76.7 | 6,173 | |
| RL + LPModel=Qwen3-32B-thinking2026.04 | 76.7 | 6,238 | |
| REDModel=Qwen3-8B-thinking2026.04 | 75.6 | 5,209 | |
| DEERModel=Qwen3-32B-thinking2026.04 | 74.4 | 7,894 | |
| REDModel=DpSk-R1-Distill-Qwen-32B2026.04 | 74.4 | 3,898 | |
| AlphaOneModel=Qwen3-8B-thinking2026.04 | 73.3 | 8,343 | |
| S-GRPOModel=Qwen3-8B-thinking2026.04 | 73.3 | 6,771 | |
| GRPOModel=DpSk-R1-Distill-Qwen-32B2026.04 | 73.3 | 8,389 | |
| REDModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 73.3 | 3,974 | |
| GRPOModel=Qwen3-8B-thinking2026.04 | 72.2 | 10,931 | |
| AlphaOneModel=DpSk-R1-Distill-Qwen-32B2026.04 | 72.2 | 8,210 | |
| AlphaOneModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 72.2 | 7,873 | |
| GRPOModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 72.2 | 8,109 | |
| DEERModel=Qwen3-8B-thinking2026.04 | 71.1 | 7,903 | |
| RL + LPModel=Qwen3-8B-thinking2026.04 | 71.1 | 6,986 | |
| RL + LPModel=DpSk-R1-Distill-Qwen-32B2026.04 | 71.1 | 5,492 | |
| Think or NotModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 71.1 | 4,005 | |
| VanillaModel=Qwen3-8B-thinking2026.04 | 70 | 11,125 | |
| VanillaModel=DpSk-R1-Distill-Qwen-32B2026.04 | 70 | 7,873 | |
| DASTModel=DpSk-R1-Distill-Qwen-32B2026.04 | 70 | 5,802 | |
| S-GRPOModel=DpSk-R1-Distill-Qwen-32B2026.04 | 70 | 4,906 | |
| DEERModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 70 | 6,829 | |
| S-GRPOModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 70 | 5,002 | |
| Think or NotModel=Qwen3-8B-thinking2026.04 | 68.9 | 5,387 | |
| DEERModel=DpSk-R1-Distill-Qwen-32B2026.04 | 68.9 | 6,461 | |
| VanillaModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 68.9 | 7,766 | |
| DASTModel=Qwen3-8B-thinking2026.04 | 67.8 | 5,964 | |
| Think or NotModel=DpSk-R1-Distill-Qwen-32B2026.04 | 67.8 | 3,993 | |
| DASTModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 67.8 | 5,115 | |
| RL + LPModel=DeepSeek-R1-Distill-Llama-70B2026.04 | 66.7 | 5,304 | |
| REDModel=DpSk-R1-Distill-Qwen-7B2026.04 | 57.8 | 4,293 | |
| GRPOModel=DpSk-R1-Distill-Qwen-7B2026.04 | 56.7 | 11,673 | |
| AlphaOneModel=DpSk-R1-Distill-Qwen-7B2026.04 | 55.6 | 8,224 | |
| DASTModel=DpSk-R1-Distill-Qwen-7B2026.04 | 55.6 | 7,258 | |
| VanillaModel=DpSk-R1-Distill-Qwen-7B2026.04 | 54.4 | 10,438 | |
| S-GRPOModel=DpSk-R1-Distill-Qwen-7B2026.04 | 54.4 | 5,094 | |
| DEERModel=DpSk-R1-Distill-Qwen-7B2026.04 | 53.3 | 7,197 | |
| Think or NotModel=DpSk-R1-Distill-Qwen-7B2026.04 | 52.2 | 4,341 | |
| RL + LPModel=DpSk-R1-Distill-Qwen-7B2026.04 | 52.2 | 5,693 | |
| AlphaOneModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 47.8 | 8,339 | |
| REDModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 47.8 | 4,507 | |
| DEERModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 46.7 | 8,001 | |
| VanillaModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 45.6 | 10,798.9 | |
| DASTModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 45.6 | 8,246 | |
| RL + LPModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 45.6 | 5,333 | |
| S-GRPOModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 45.6 | 4,809 | |
| Think or NotModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 44.4 | 6,761 | |
| GRPOModel=DeepSeek-R1-Distill-Llama-8B2026.04 | 44.4 | 11,312 |