Reasoning on HLE
18.03Pass@1Free()LM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Free()LMBackbone=Qwen3-235B2026.02 | 18.03 | 15.8 | |
| GeminiBackbone=Qwen3-235B2026.02 | 16.59 | 18.7 | |
| VanillaBackbone=Qwen3-235B2026.02 | 16.13 | 22.5 | |
| No TrainBackbone=Qwen3-235B2026.02 | 15.2 | 18.2 | |
| Free()LMBackbone=Qwen3-30B2026.02 | 9.82 | 11.3 | |
| No TrainBackbone=Qwen3-30B2026.02 | 9.78 | 12.7 | |
| GeminiBackbone=Qwen3-30B2026.02 | 9.59 | 12.8 | |
| VanillaBackbone=Qwen3-30B2026.02 | 8.99 | 13.9 | |
| No TrainBackbone=Qwen3-8B2026.02 | 5.89 | 12.3 | |
| Free()LMBackbone=Qwen3-8B2026.02 | 5.38 | 9.9 | |
| GeminiBackbone=Qwen3-8B2026.02 | 5.1 | 11.8 | |
| TCBackbone=Qwen3-8B2026.02 | 4.96 | 15.3 | |
| VanillaBackbone=Qwen3-8B2026.02 | 4.59 | 13.3 | |
| H2OBackbone=Qwen3-8B2026.02 | 4.39 | 15.3 |