List your company

Mathematical Reasoning on AMC23 (Maj@16, Avg@16)

95Maj@16GRPO
1.425.75074.3Jun 7, 2026

Evaluation results

MethodLinks
GRPOModel=Qwen3-4B-Base2026.069587.2
PAECModel=Qwen3-4B-Base2026.0692.587.8
BaseModel=Qwen3-4B-Base2026.069082.9
Global Entropy RegModel=Qwen3-4B-Base2026.069082.9
PAECModel=Llama3.1-8B2026.061510.4
Global Entropy RegModel=Llama3.1-8B2026.06108.4
GRPOModel=Llama3.1-8B2026.067.57.2
BaseModel=Llama3.1-8B2026.0653