List your company

Mathematical Reasoning on CNK12 (test)

0Cost ($)SLM
-0.022360.128570.27950.43043Jan 29, 2026

Evaluation results

MethodLinks
SLMStrategy Type=Baseline2026.01053.8——
Oracle ShepherdingStrategy Type=Cascading2026.010.30684.445.21.83
Reactive ShepherdingStrategy Type=Cascading2026.010.3247642.11.25
FrugalGPTStrategy Type=Cascading2026.010.39176.130.11.04
Proactive ShepherdingStrategy Type=Routing2026.010.45577.918.50.97
ABCStrategy Type=Cascading2026.010.4781.315.81.07
GraphRouterStrategy Type=Routing2026.010.51381.78.20.99
RouteLLMStrategy Type=Routing2026.010.53382.64.650.99
LLMStrategy Type=Baseline2026.010.55984.4——