List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME 25 (Speedup, tau)

5.97SpeedupDOMINO
1.69562.80533.9155.0247May 28, 2026
Updated 5mo ago

Evaluation results

MethodLinks
DOMINOTemperature=0, Model=Qwen3-4B, Draft size=162026.055.977.49
DFlashTemperature=0, Model=Qwen3-4B, Draft size=162026.055.877.47
DOMINOTemperature=0, Model=Qwen3-8B, Draft size=162026.055.857.41
DFlashTemperature=0, Model=Qwen3-8B, Draft size=162026.055.677.13
DFlashTemperature=1, Model=Qwen3-4B, Draft size=162026.053.795.05
DOMINOTemperature=1, Model=Qwen3-8B, Draft size=162026.053.784.92
DOMINOTemperature=1, Model=Qwen3-4B, Draft size=162026.053.754.83
DFlashTemperature=1, Model=Qwen3-8B, Draft size=162026.053.724.77
EAGLE-3Temperature=0, Model=Qwen3-8B, Draft size=602026.052.413.59
EAGLE-3Temperature=0, Model=Qwen3-4B, Draft size=602026.052.363.53
DARTTemperature=0, Model=Qwen3-4B, Draft size=602026.052.132.59
DARTTemperature=1, Model=Qwen3-4B, Draft size=602026.052.132.59
DARTTemperature=1, Model=Qwen3-8B, Draft size=602026.052.132.59
DARTTemperature=0, Model=Qwen3-8B, Draft size=602026.052.112.59
EAGLE-3Temperature=0, Model=Qwen3-4B, Draft size=162026.052.083.1
EAGLE-3Temperature=1, Model=Qwen3-4B, Draft size=602026.052.083.1
EAGLE-3Temperature=0, Model=Qwen3-8B, Draft size=162026.052.073.07
EAGLE-3Temperature=1, Model=Qwen3-8B, Draft size=602026.052.053.17
EAGLE-3Temperature=1, Model=Qwen3-4B, Draft size=162026.051.862.8
EAGLE-3Temperature=1, Model=Qwen3-8B, Draft size=162026.051.862.8