Mathematical Reasoning on HMMT25
83.9Pass@1RSE
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| RSEBackbone=Qwen3-30B-A3B-Thinking-2507, Inference Strategy=Recycling Search Experience, Iteration=It32026.01 | 83.9 | — | — | — | |
| RSEBackbone=Qwen3-4B-Thinking-2507, Inference Strategy=Recycling Search Experience, Iteration=It32026.01 | 73.5 | — | — | — | |
| PaCoReBackbone=Qwen3-4B-Thinking-2507, Inference Strategy=PaCoRe, Iteration=It12026.01 | 69.6 | — | — | — | |
| BaseBackbone=Qwen3-30B-A3B-Thinking-2507, Inference Strategy=Standard Sampling, Iteration=It02026.01 | 69 | — | — | — | |
| RSEBackbone=Phi-4-Reasoning, Inference Strategy=Recycling Search Experience, Iteration=It32026.01 | 67.5 | — | — | — | |
| Pass@1Model=DeepSeek-8B, Number of repeated trials=162026.02 | 60.42 | — | — | — | |
| Pass@1Model=Qwen3-4B, Number of repeated trials=162026.02 | 55.42 | — | — | — | |
| BaseBackbone=Qwen3-4B-Thinking-2507, Inference Strategy=Standard Sampling, Iteration=It02026.01 | 54 | — | — | — | |
| BaseBackbone=Phi-4-Reasoning, Inference Strategy=Standard Sampling, Iteration=It02026.01 | 43.9 | — | — | — | |
| Pass@1Model=DeepSeek-1.5B, Number of repeated trials=162026.02 | 12.08 | — | — | — | |
| Chronos@128Model=DeepSeek-1.5B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | — | 27.92 | |
| Chronos@128Model=Qwen3-4B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | — | 74.38 | |
| Chronos@128Model=DeepSeek-8B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | — | 76.04 | |
| DeepConf@128Model=DeepSeek-1.5B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | 25.21 | — | |
| DeepConf@128Model=Qwen3-4B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | 67.08 | — | |
| DeepConf@128Model=DeepSeek-8B, Number of repeated trials=16, Trajectories=1282026.02 | — | — | 73.13 | — | |
| Maj@128Model=DeepSeek-1.5B, Number of repeated trials=16, Trajectories=1282026.02 | — | 16.67 | — | — | |
| Maj@128Model=Qwen3-4B, Number of repeated trials=16, Trajectories=1282026.02 | — | 60.62 | — | — | |
| Maj@128Model=DeepSeek-8B, Number of repeated trials=16, Trajectories=1282026.02 | — | 68.75 | — | — |