List your company

Mathematical Reasoning on Theorem (test)

32.2AccuracyDeepSeekMath-DART
18.36821.95925.5529.141Apr 13, 2026

Evaluation results

MethodLinks
DeepSeekMath-DARTBackbone=DeepSeekMath-7B, #Samples=590K2026.0432.2
DeepSeekMath-CRPS-60KBackbone=DeepSeekMath-7B, #Samples=60K2026.0431.5
SIGMA-60KBackbone=DeepSeekMath-7B, #Samples=60K2026.0429.3
DeepSeekMath-CRPS-30KBackbone=DeepSeekMath-7B, #Samples=30K2026.0429.2
DeepSeekMath-InstructBackbone=DeepSeekMath-7B, #Samples=780K2026.0428.1
DeepSeekMath-CRPS-15KBackbone=DeepSeekMath-7B, #Samples=15K2026.0427.8
DeepSeekMath-DARTBackbone=DeepSeekMath-7B, #Samples=60K2026.0427.4
DeepSeekMath-RFTBackbone=DeepSeekMath-7B, #Samples=590K2026.0427.2
SIGMA-30KBackbone=DeepSeekMath-7B, #Samples=30K2026.0426.6
MathFusionBackbone=DeepSeekMath-7B, #Samples=60K2026.0424.6
DeepSeekMath-MMIQCBackbone=DeepSeekMath-7B, #Samples=2.3M2026.0423.4
MathFusion (Sequential)Backbone=DeepSeekMath-7B, #Samples=30K2026.0422.8
DeepSeekMath-MetaMathBackbone=DeepSeekMath-7B, #Samples=60K2026.0418.9