List your company
ResearchFormerly WizWand

Mathematical Reasoning on Math Word

24.6AccuracyLIDAS
2.02167.883313.74519.6067Dec 9, 2025
Updated 8mo ago

Evaluation results

MethodLinks
LIDASModel Size=1.7B, Cooldown Protocol=Math, Five-shot prompting=true2025.1224.6
MIDASModel Size=1.7B, Cooldown Protocol=Math, Five-shot prompting=true2025.1224.01
BaselineModel Size=1.7B, Cooldown Protocol=Math, Five-shot prompting=true2025.1223.28
LIDASModel Size=1.7B, Cooldown Protocol=Standard, Five-shot prompting=true2025.1218.59
LN-ScalingModel Size=1.7B, Cooldown Protocol=Math, Five-shot prompting=true2025.1217.84
MIDASModel Size=1.7B, Cooldown Protocol=Standard, Five-shot prompting=true2025.1216.07
BaselineModel Size=1.7B, Cooldown Protocol=Standard, Five-shot prompting=true2025.1213.75
MIDASModel Size=360M, Cooldown Protocol=Math, Five-shot prompting=true2025.1213.43
LIDASModel Size=360M, Cooldown Protocol=Math, Five-shot prompting=true2025.1212.3
LN-ScalingModel Size=1.7B, Cooldown Protocol=Standard, Five-shot prompting=true2025.1211
LN-ScalingModel Size=360M, Cooldown Protocol=Math, Five-shot prompting=true2025.128.45
BaselineModel Size=360M, Cooldown Protocol=Math, Five-shot prompting=true2025.128.1
MIDASModel Size=360M, Cooldown Protocol=Standard, Five-shot prompting=true2025.124.39
LIDASModel Size=360M, Cooldown Protocol=Standard, Five-shot prompting=true2025.124.36
BaselineModel Size=360M, Cooldown Protocol=Standard, Five-shot prompting=true2025.123.69
LN-ScalingModel Size=360M, Cooldown Protocol=Standard, Five-shot prompting=true2025.122.89