List your company

Math and Knowledge Reasoning on AIME 2024 (Avg)

86.3Average ScoreRL initialization model
85.530485.730285.9386.1298May 13, 2026

Evaluation results

MethodLinks
RL initialization modelSampling protocol=@322026.0586.3
RDPOSampling protocol=@322026.0585.73
GRPOSampling protocol=@322026.0585.56