List your company

Mathematical Reasoning on MATH (Acc@8 and Relative Change)

57.29Acc@8GRPO
-2.291613.176728.64544.1133Jul 2, 2026

Evaluation results

MethodLinks
GRPOStage=Last2026.0757.2975
SDPO0Stage=Current2026.0756.4272
GRPOStage=Current2026.0744.6737
SDPOCoT0Stage=Current2026.0738.1317
SDPO0Stage=Last2026.0734.385
SDPO5%Stage=Current2026.0734.385
BaselineStage=Baseline2026.0732.71—
SDPOCoT0Stage=Last2026.0714.1757
SDPO5%Stage=Last2026.070100