Longterm Wiki has not been actively maintained since May 2026. Some data is still updated occasionally, but please don't rely on this site being current or accurate.

WinoGrande

Reasoning

A large-scale commonsense reasoning benchmark with 44,000 Winograd-schema-style problems, using adversarial filtering to reduce annotation artifacts.

Models Tested

1

Best Score

81.6

Median Score

81.6

Scoring: accuracy

Introduced: 2019-07

Maintainer: AI2

Leaderboard (1 model)

#	Model	Developer	Score
🥇	GPT-3.5 Turbo	OpenAI	81.6