List your company
ResearchFormerly WizWand

Math Word Problem Solving on GSM8K (test)

89.3AccuracyMO-CAPO
42.08454.34266.678.858May 15, 2026
Updated 5mo ago

Evaluation results

MethodLinks
MO-CAPOModel=GPT-OSS-120B, Token Budget=7.5M2026.0589.3
CAPOModel=GPT-OSS-120B, Token Budget=7.5M2026.0587.1
GEPAModel=GPT-OSS-120B, Token Budget=7.5M2026.0585.7
NSGA-II-POModel=GPT-OSS-120B, Token Budget=7.5M2026.0585.2
CAPOModel=Mistral-3.2-24B, Token Budget=7.5M2026.0575.2
MO-CAPOModel=Mistral-3.2-24B, Token Budget=7.5M2026.0575.1
NSGA-II-POModel=Mistral-3.2-24B, Token Budget=7.5M2026.0574.3
CAPOModel=Qwen3-30B, Token Budget=7.5M2026.0566.5
MO-CAPOModel=Qwen3-30B, Token Budget=7.5M2026.0566.2
GEPAModel=Qwen3-30B, Token Budget=7.5M2026.0565.9
NSGA-II-POModel=Qwen3-30B, Token Budget=7.5M2026.0565.7
EvoPromptGAModel=GPT-OSS-120B, Token Budget=7.5M2026.0558.4
GEPAModel=Mistral-3.2-24B, Token Budget=7.5M2026.0557.1
InitialModel=Qwen3-30B, Token Budget=7.5M2026.0554
EvoPromptGAModel=Qwen3-30B, Token Budget=7.5M2026.0553.7
InitialModel=GPT-OSS-120B, Token Budget=7.5M2026.0549.5
InitialModel=Mistral-3.2-24B, Token Budget=7.5M2026.0546.3
EvoPromptGAModel=Mistral-3.2-24B, Token Budget=7.5M2026.0543.9