WinoGrande

WinoGrande - large-scale commonsense reasoning with Winograd schema-like problems. 5-shot.

Category

Reasoning

Max Score

100.0

Score Type

percent

Active

Yes

Model Scores

Scores for WinoGrande

Model Score Date Verified
Gemma 3 1B PT
58.20
27.02.2025 Verified
Gemma 3 4B PT
31.55%
27.02.2025 Verified
Gemma 3 12B PT
78.16%
27.02.2025 Verified
Gemma 3 27B PT
100.00%
27.02.2025 Verified