Reasoning on AIME 2025 (Speedup, Mean acceptance length τ)
7.23SpeedupDFlash+DDTree
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DFlash+DDTreeModel=Qwen3-4B, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 7.23 | 10.23 | |
| DFlash+DDTreeModel=Qwen3-8B, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 6.99 | 9.86 | |
| DFlash+DDTreeModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 5.88 | 7.63 | |
| DFlashModel=Qwen3-4B, Temperature=0.02026.04 | 5.33 | 7.37 | |
| DFlashModel=Qwen3-8B, Temperature=0.02026.04 | 5.32 | 7.39 | |
| DFlash+DDTreeModel=Qwen3-8B, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 5.25 | 7.71 | |
| DFlash+DDTreeModel=Qwen3-4B, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 5.08 | 7.22 | |
| DFlash+DDTreeModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 4.98 | 6.62 | |
| DFlashModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=0.02026.04 | 3.98 | 5.13 | |
| DFlashModel=Qwen3-4B, Temperature=1.02026.04 | 3.38 | 4.79 | |
| DFlashModel=Qwen3-8B, Temperature=1.02026.04 | 3.36 | 4.79 | |
| DFlashModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=1.02026.04 | 3.17 | 4.26 |