Mathematical Reasoning on AIME 24 (AUCOAA)
81.8AUCOAAFormat-Adaptive-Answer
Evaluation Results
| Method | Links | |
|---|---|---|
| Format-Adaptive-AnswerBackbone=Qwen3-8B2026.01 | 81.8 | |
| Normalized-LengthBackbone=Qwen3-8B2026.01 | 77.4 | |
| Adaptive-AnswerBackbone=Qwen3-8B2026.01 | 75.8 | |
| Hard-Length 8k → 4kBackbone=Qwen3-8B2026.01 | 75.1 | |
| TWYNBackbone=Qwen3-8B2026.01 | 74.5 | |
| SFTBackbone=Qwen3-8B2026.01 | 73.7 | |
| Hard-Length 8kBackbone=Qwen3-8B2026.01 | 73.3 | |
| Soft-LengthBackbone=Qwen3-8B2026.01 | 72.4 | |
| Hard-Length 16kBackbone=Qwen3-8B2026.01 | 70.7 | |
| Base modelBackbone=Qwen3-8B2026.01 | 68.6 | |
| No-ThinkingBackbone=Qwen3-8B2026.01 | 21.9 |