List your company

Mathematical Reasoning on GSM8K (AUC %)

85.8AUC (%)CF
75.19277.94680.783.454Feb 10, 2026

Evaluation results

MethodLinks
CFModel=Qwen2.5-3B2026.0285.8
DAPOModel=Qwen2.5-3B2026.0285.3
RandomModel=Qwen2.5-3B2026.0285
InvertedModel=Qwen2.5-3B2026.0284.8
CFModel=Qwen3-1.7B2026.0283.2
DAPOModel=Qwen3-1.7B2026.0282.5
RandomModel=Qwen3-1.7B2026.0282.1
InvertedModel=Qwen3-1.7B2026.0281
CFModel=Llama3.2-3B2026.0277.3
DAPOModel=Llama3.2-3B2026.0276.6
RandomModel=Llama3.2-3B2026.0275.9
InvertedModel=Llama3.2-3B2026.0275.6