Mathematical Reasoning on AIME 26
86.7ScoreGpt-oss-20b-high
Evaluation Results
| Method | Links | |
|---|---|---|
| Gpt-oss-20b-highGroup=Larger, Generation Configuration=Standardized2026.03 | 86.7 | |
| Qwen3-14BGroup=Larger, Generation Configuration=Standardized2026.03 | 86.7 | |
| Qwen3-30B-A3BGroup=Larger, Generation Configuration=Standardized2026.03 | 86.7 | |
| EngGPT2-16B-A3BGroup=Comparable, Generation Configuration=Standardized2026.03 | 70 | |
| p1Backbone=Qwen3-4B-Instruct-2507, S=[1, 23], M=16, Training Reward Imp.=0.212026.04 | 62.24 | |
| p1Backbone=Qwen3-4B-Instruct-2507, S=[17, 27], M=16, Training Reward Imp.=0.092026.04 | 57.24 | |
| p1Backbone=Qwen3-4B-Instruct-2507, S=[4, 5, 17, 20], M=8, Training Reward Imp.=0.042026.04 | 55.68 | |
| p1Backbone=Qwen3-4B-Instruct-2507, S=[23], M=32, Training Reward Imp.=0.072026.04 | 55.21 | |
| RLBackbone=Qwen3-4B-Instruct-2507, S=all 30, M=1, Training Reward Imp.=02026.04 | 54.58 | |
| baseBackbone=Qwen3-4B-Instruct-2507, S=/, M=/, Training Reward Imp.=/2026.04 | 54.38 | |
| GEPABackbone=Qwen3-4B-Instruct-2507, S=15/15, M=/, Training Reward Imp.=/2026.04 | 54.22 | |
| GEPABackbone=Qwen3-4B-Instruct-2507, S=28/[1,23], M=/, Training Reward Imp.=/2026.04 | 53.91 | |
| GEPABackbone=Qwen3-4B-Instruct-2507, S=[1,23]/28, M=/, Training Reward Imp.=/2026.04 | 53.49 | |
| p1Backbone=Qwen3-4B-Instruct-2507, S=[25], M=32, Training Reward Imp.=0.072026.04 | 51.25 | |
| p1Backbone=Qwen3-1.7B, S=[5, 26], M=16, Training Reward Imp.=0.132026.04 | 37.6 | |
| gemma-3-12b-itGroup=Larger, Generation Configuration=Standardized2026.03 | 36.7 | |
| p1Backbone=Qwen3-1.7B, S=[10, 11], M=16, Training Reward Imp.=0.142026.04 | 36.15 | |
| p1Backbone=Qwen3-1.7B, S=[5, 14, 20, 26], M=8, Training Reward Imp.=0.052026.04 | 35.99 | |
| RLBackbone=Qwen3-1.7B, S=all 30, M=1, Training Reward Imp.=02026.04 | 35.89 | |
| p1Backbone=Qwen3-1.7B, S=[5], M=32, Training Reward Imp.=0.252026.04 | 35.63 | |
| baseBackbone=Qwen3-1.7B, S=/, M=/, Training Reward Imp.=/2026.04 | 35.57 | |
| GEPABackbone=Qwen3-1.7B, S=28/[5,26], M=/, Training Reward Imp.=/2026.04 | 35.52 | |
| GEPABackbone=Qwen3-1.7B, S=[5,26]/28, M=/, Training Reward Imp.=/2026.04 | 35.26 | |
| GEPABackbone=Qwen3-1.7B, S=15/15, M=/, Training Reward Imp.=/2026.04 | 35.1 | |
| p1Backbone=Qwen3-1.7B, S=[26], M=32, Training Reward Imp.=0.292026.04 | 33.23 | |
| Entropy-Cut MHModel=Qwen2.5-Math-7B2026.05 | 13.1 | |
| SMCModel=Qwen2.5-Math-7B2026.05 | 11.8 | |
| TMCModel=Qwen2.5-Math-7B2026.05 | 10.6 | |
| Entropy-Cut MHModel=Qwen3-8B-Base2026.05 | 10.3 | |
| SMCModel=Qwen3-8B-Base2026.05 | 10.1 | |
| Moonlight-16B-A3B-InstructGroup=Comparable, Generation Configuration=Standardized2026.03 | 10 | |
| TMCModel=Qwen3-8B-Base2026.05 | 9.8 | |
| Uniform-Cut MHModel=Qwen2.5-Math-7B2026.05 | 9.5 | |
| Uniform-Cut MHModel=Qwen3-8B-Base2026.05 | 9.5 | |
| TMCModel=Qwen2.5-7B2026.05 | 9.4 | |
| Entropy-Cut MHModel=Qwen2.5-7B2026.05 | 9.4 | |
| Entropy-Cut MHModel=Phi-4-mini-instruct2026.05 | 8.7 | |
| SMCModel=Phi-4-mini-instruct2026.05 | 8.6 | |
| Uniform-Cut MHModel=Qwen2.5-7B2026.05 | 8.2 | |
| Uniform-Cut MHModel=Phi-4-mini-instruct2026.05 | 8 | |
| SMCModel=Qwen2.5-7B2026.05 | 7.4 | |
| TMCModel=Phi-4-mini-instruct2026.05 | 6.9 | |
| Low-TemperatureModel=Phi-4-mini-instruct2026.05 | 6.8 | |
| Low-TemperatureModel=Qwen2.5-Math-7B2026.05 | 6.4 | |
| Low-TemperatureModel=Qwen3-8B-Base2026.05 | 6.4 | |
| Low-TemperatureModel=Qwen2.5-7B2026.05 | 5.7 | |
| StandardModel=Qwen2.5-Math-7B2026.05 | 4.8 | |
| StandardModel=Phi-4-mini-instruct2026.05 | 3.5 | |
| StandardModel=Qwen3-8B-Base2026.05 | 3.2 | |
| Entropy-Cut MHModel=Phi-3.5-mini-instruct2026.05 | 3.1 | |
| TMCModel=Phi-3.5-mini-instruct2026.05 | 3 | |
| Uniform-Cut MHModel=Phi-3.5-mini-instruct2026.05 | 2.9 | |
| StandardModel=Qwen2.5-7B2026.05 | 2 | |
| SMCModel=Phi-3.5-mini-instruct2026.05 | 1.7 | |
| Low-TemperatureModel=Phi-3.5-mini-instruct2026.05 | 1.6 | |
| StandardModel=Phi-3.5-mini-instruct2026.05 | 0.6 | |
| Llama-3.1-8B-InstructGroup=Comparable, Generation Configuration=Standardized2026.03 | 0.1 | |
| gemma-2-9b-itGroup=Comparable, Generation Configuration=Standardized2026.03 | 0 |