List your company

Reasoning on Fables (test)

0.719AUROCTRACED
0.717960.718230.71850.71877Mar 11, 2026

Evaluation results

MethodLinks
TRACEDBase Model=DeepSeek-R12026.030.7190.034
LR ProbeBase Model=DeepSeek-R1, Method Status=Runner-up Baseline2026.030.718—