List your company

Commonsense Reasoning on Wino-X DE

79.3Normalized Log AccuracyHATified-SFT
75.97276.83677.778.564Mar 16, 2026

Evaluation results

MethodLinks
HATified-SFTMode=SFT, Shots=5-shot2026.0379.3
Llama-InstructMode=SFT, Shots=5-shot2026.0376.1