List your company

Mathematical Reasoning on AMC 2023 (pass@3)

85.79pass@3HDPO
58.032465.238772.44579.6513Jun 2, 2026

Evaluation results

MethodLinks
HDPOBackbone=Qwen2.5-Math-7B, Number of samples=162026.0685.79
HDPOBackbone=DS-R1-Llama-8B, Number of samples=162026.0684.17
HDPOBackbone=Qwen3-4B, Number of samples=162026.0683.26
Critique-GRPOBackbone=8B, Number of samples=162026.0681.88
GRPOBackbone=Qwen2.5-Math-7B, Number of samples=162026.0680.92
LUFFYBackbone=Qwen-Math-7B, Number of samples=162026.0679.68
GRPOBackbone=DS-R1-Llama-8B, Number of samples=162026.0676.62
SimpleRLBackbone=Qwen-2.5-7B, Number of samples=162026.0674.27
EDGE-GRPOBackbone=Qwen-7B, Number of samples=162026.0673.79
OatBackbone=Qwen2.5-Math-7B, Number of samples=162026.0673.59
GRPOBackbone=Qwen3-4B, Number of samples=162026.0673.32
Base ModelBackbone=Qwen2.5-Math-7B, Number of samples=162026.0671.08
Base ModelBackbone=DS-R1-Llama-8B, Number of samples=162026.0663.78
Base ModelBackbone=Qwen3-4B, Number of samples=162026.0659.1