List your company

Mathematical Reasoning on GSM8k (Pass@1, Pass@16, Major@16)

93.9Pass@1DAPO
81.83684.96888.191.232May 18, 2026

Evaluation results

MethodLinks
DAPOBackbone=Qwen2.5-14B2026.0593.997.995.5
CorR-POBackbone=Qwen2.5-14B2026.0593.698.393.9
GSPOBackbone=Qwen2.5-14B2026.0592.697.894.2
GRPOBackbone=Qwen2.5-14B2026.0591.798.393.9
Dr.GRPOBackbone=Qwen2.5-14B2026.0591.497.594.1
BaseBackbone=Qwen2.5-14B2026.0582.397.390.2