Reasoning on BeyondAIME
70.38Pass@1Gemini
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GeminiBackbone=Qwen3-235B2026.02 | 70.38 | 26.5 | |
| Free()LMBackbone=Qwen3-235B2026.02 | 69.75 | 21.3 | |
| VanillaBackbone=Qwen3-235B2026.02 | 69 | 31 | |
| No TrainBackbone=Qwen3-235B2026.02 | 68.63 | 22.9 | |
| Free()LMBackbone=Qwen3-30B2026.02 | 62.25 | 18.2 | |
| GeminiBackbone=Qwen3-30B2026.02 | 59.75 | 19.7 | |
| No TrainBackbone=Qwen3-30B2026.02 | 59.25 | 20.8 | |
| VanillaBackbone=Qwen3-30B2026.02 | 56.5 | 22.5 | |
| Free()LMBackbone=Qwen3-8B2026.02 | 45.88 | 15.4 | |
| GeminiBackbone=Qwen3-8B2026.02 | 43.63 | 17.6 | |
| No TrainBackbone=Qwen3-8B2026.02 | 43.38 | 20.6 | |
| VanillaBackbone=Qwen3-8B2026.02 | 42.38 | 19.5 | |
| TCBackbone=Qwen3-8B2026.02 | 42 | 19.7 | |
| H2OBackbone=Qwen3-8B2026.02 | 35 | 20.5 |