List your company

Common Sense Reasoning on Wino-X DE

76.3Normalized Log AccuracyHATified
65.27668.1387173.862Mar 16, 2026

Evaluation results

MethodLinks
HATifiedtraining=SFT, shots=5-shot2026.0376.35.63
Tülutraining=SFT, shots=5-shot2026.0365.73.67