Mathematical Reasoning on MMLU algebra
97Pass@1CoD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CoDBackbone=QwQ-32B2026.06 | 97 | 1,345 | |
| DyConBackbone=QwQ-32B2026.06 | 97 | 1,266 | |
| NothinkingBackbone=QwQ-32B2026.06 | 96 | 1,431 | |
| DEERBackbone=QwQ-32B2026.06 | 96 | 1,395 | |
| BaselineBackbone=Qwen3-14B2026.06 | 96 | 2,545 | |
| DyConBackbone=Qwen3-14B2026.06 | 96 | 2,073 | |
| CoDBackbone=Qwen3-4B-Thinking-25072026.06 | 95 | 3,209 | |
| NoWaitBackbone=Qwen3-4B-Thinking-25072026.06 | 95 | 2,068 | |
| DyConBackbone=Qwen3-4B-Thinking-25072026.06 | 95 | 2,122 | |
| BaselineBackbone=QwQ-32B2026.06 | 95 | 2,133 | |
| SEALBackbone=QwQ-32B2026.06 | 95 | 1,541 | |
| NoWaitBackbone=QwQ-32B2026.06 | 95 | 1,302 | |
| BaselineBackbone=Qwen3-4B-Thinking-25072026.06 | 94 | 3,496 | |
| NothinkingBackbone=Qwen3-4B-Thinking-25072026.06 | 94 | 2,331 | |
| CoDBackbone=Qwen3-14B2026.06 | 94 | 1,381 | |
| NoWaitBackbone=Qwen3-14B2026.06 | 93 | 1,729 | |
| DEERBackbone=Qwen3-4B-Thinking-25072026.06 | 92 | 1,945 | |
| DyConBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 91 | 1,488 | |
| Dynasor-CoTBackbone=Qwen3-14B2026.06 | 91 | 1,733 | |
| BaselineBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 90 | 2,387 | |
| Controlling Thinking SpeedBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 90 | 1,719 | |
| NoWaitBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 89 | 1,347 | |
| ThinkpilotBackbone=Qwen3-14B2026.06 | 88 | 538 | |
| DEERBackbone=Qwen3-14B2026.06 | 87 | 1,380 | |
| CoDBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 85 | 1,091 | |
| NothinkingBackbone=Qwen3-14B2026.06 | 84 | 547 | |
| ThinkpilotBackbone=Qwen3-4B-Thinking-25072026.06 | 83 | 1,306 | |
| SEALBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 80 | 1,507 | |
| DEERBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 79 | 1,493 | |
| NothinkingBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 74 | 760 | |
| ThinkpilotBackbone=DeepSeek-R1-Distill-Qwen-7B2026.06 | 74 | 705 |