List your company
ResearchFormerly WizWand

Mathematical Reasoning on GSM8K (pass@1, standard deviation)

88.2Accuracy (pass@1)DeepSeekMath-7B-RL
62.82469.4127682.588May 25, 2025
Updated 5mo ago

Evaluation results

MethodLinks
DeepSeekMath-7B-RLTrained Model=7B, Zero-shot=true, Temperature=0.62025.0588.2—
Mathstral-7B-v0.1Trained Model=7B, Zero-shot=true, Temperature=0.62025.0584.9—
UniRBackbone=Qwen2.5-3B, Trained Model=1.5B, Zero-shot=true, Temperature=0.62025.0584.5—
Internlm2-math-plus-7BTrained Model=7B, Zero-shot=true, Temperature=0.62025.0584—
GRPO FullBackbone=Qwen2.5-3B, Trained Model=3B, Zero-shot=true, Temperature=0.62025.0581.3—
UniRBackbone=Llama3.2-3B, Trained Model=1B, Zero-shot=true, Temperature=0.62025.0577.5—
Backbone + 1.5BBackbone=Qwen2.5-3B, Zero-shot=true, Temperature=0.62025.0575.6—
NuminaMath-7B-CoTTrained Model=7B, Zero-shot=true, Temperature=0.62025.0575.4—
GRPO LoRABackbone=Qwen2.5-3B, Trained Model=3B, Zero-shot=true, Temperature=0.62025.0575—
Backbone onlyBackbone=Qwen2.5-3B, Zero-shot=true, Temperature=0.62025.0574.4—
GRPO LoRABackbone=Llama3.2-3B, Trained Model=3B, Zero-shot=true, Temperature=0.62025.0565.8—
Backbone onlyBackbone=Llama3.2-3B, Zero-shot=true, Temperature=0.62025.0565.6—
GRPO FullBackbone=Llama3.2-3B, Trained Model=3B, Zero-shot=true, Temperature=0.62025.0565.1—
Backbone + 1BBackbone=Llama3.2-3B, Zero-shot=true, Temperature=0.62025.0563.8—