List your company

Mathematical Reasoning on MATH 500 (Acc., Cost)

99.2AccuracyQwen3-235B
92.9694.5896.297.82Jun 28, 2026

Evaluation results

MethodLinks
Qwen3-235BMethod Category=Heuristic, Parameters=235B2026.0699.22.5881
ENTROROUTERMethod Category=Routing Architecture2026.0698.40.5238
Qwen3-30BMethod Category=Heuristic, Parameters=30B2026.0698.21.2123
FORCMethod Category=Existing Routing Architecture2026.0698.21.7675
RandomMethod Category=Heuristic2026.06981.279
RouteLLMMethod Category=Existing Routing Architecture2026.0697.60.1042
Qwen3-4BMethod Category=Heuristic, Parameters=4B2026.0697.40.0907
Router-R1Method Category=Existing Routing Architecture2026.0693.40.0018
xRouterMethod Category=Existing Routing Architecture2026.0693.20.0009