List your company

Mathematical Reasoning on GSM8K (Accuracy, Reusability, Verifiability)

94AccuracyDeepSeek-R1
77.3681.688690.32Feb 19, 2026

Evaluation results

MethodLinks
DeepSeek-R12026.02946780
Gemma32026.02936679
Phi4-Reasoning2026.02818488
Llama3.12026.02787371