List your company

Reasoning on GSM8K (test) using AUROC/p-value

80.6AUROCTRACED
79.97680.13880.380.462Mar 11, 2026

Evaluation results

MethodLinks
TRACEDBase Model=DeepSeek-R12026.0380.60.032
SAPLMABase Model=DeepSeek-R1, Method Status=Runner-up Baseline2026.0380—