Reasoning on MATH-500 (Speedup, Mean acceptance length (τ))
7.52SpeedupDFlash+DDTree
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DFlash+DDTreeModel=Qwen3-8B, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 7.52 | 10.73 | |
| DFlash+DDTreeModel=Qwen3-4B, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 7.5 | 10.71 | |
| DFlash+DDTreeModel=Qwen3-4B, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 6.6 | 9.61 | |
| DFlash+DDTreeModel=Qwen3-8B, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 6.59 | 9.54 | |
| DFlash+DDTreeModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=0.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 6.21 | 8.1 | |
| DFlash+DDTreeModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=1.0, Node Budget=Best from {16, 32, 64, 128, 256, 512, 1024}2026.04 | 5.91 | 7.76 | |
| DFlashModel=Qwen3-8B, Temperature=0.02026.04 | 5.56 | 7.79 | |
| DFlashModel=Qwen3-4B, Temperature=0.02026.04 | 5.54 | 7.72 | |
| DFlashModel=Qwen3-4B, Temperature=1.02026.04 | 4.65 | 6.6 | |
| DFlashModel=Qwen3-8B, Temperature=1.02026.04 | 4.56 | 6.46 | |
| DFlashModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=0.02026.04 | 4.29 | 5.58 | |
| DFlashModel=Qwen3-Coder-30B-A3B-Instruct, Temperature=1.02026.04 | 4.01 | 5.32 |