Mathematical Reasoning on AIME 25 (Accepted length (τ))
5.12Accepted Length (τ)DSpark
Evaluation Results
| Method | Links | |
|---|---|---|
| DSparktarget_model=Gemma4-12B2026.07 | 5.12 | |
| DSparktarget_model=Qwen3-8B2026.07 | 5.01 | |
| DSparktarget_model=Qwen3-14B2026.07 | 4.94 | |
| DSparktarget_model=Qwen3-4B2026.07 | 4.89 | |
| Eagle3target_model=Gemma4-12B2026.07 | 4.83 | |
| DFlashtarget_model=Gemma4-12B2026.07 | 4.22 | |
| DFlashtarget_model=Qwen3-4B2026.07 | 4.15 | |
| DFlashtarget_model=Qwen3-8B2026.07 | 4.07 | |
| DFlashtarget_model=Qwen3-14B2026.07 | 3.98 | |
| Eagle3target_model=Qwen3-4B2026.07 | 3.92 | |
| Eagle3target_model=Qwen3-8B2026.07 | 3.91 | |
| Eagle3target_model=Qwen3-14B2026.07 | 3.71 |