List your company

Mathematical Reasoning on HMMT 02 26 (pass@1, pass@64)

11.9Pass@1CurveRL
0.6683.5846.59.416May 23, 2026

Evaluation results

MethodLinks
CurveRLBackbone=Qwen3-4B-Base2026.0511.940.9
MaxRLBackbone=Qwen3-4B-Base2026.0511.833.9
GRPOBackbone=Qwen3-4B-Base2026.059.927.2
GRPOBackbone=Qwen3-1.7B-Base2026.053.815.5
MaxRLBackbone=Qwen3-1.7B-Base2026.053.217.7
CurveRLBackbone=Qwen3-1.7B-Base2026.05321
BaseBackbone=Qwen3-4B-Base2026.05320.6
BaseBackbone=Qwen3-1.7B-Base2026.051.114.1