List your company

Mathematical Reasoning on BRUMO 2025 (Majority@2048)

23.3Majority@2048CurveRL
12.915.618.321May 23, 2026

Evaluation results

MethodLinks
CurveRLBackbone=Qwen3-1.7B-Base2026.0523.3
GRPOBackbone=Qwen3-1.7B-Base2026.0516.7
MaxRLBackbone=Qwen3-1.7B-Base2026.0516.7
BaseBackbone=Qwen3-1.7B-Base2026.0513.3