Commonsense Reasoning
Reasoning about everyday situations (CommonsenseQA, HellaSwag).
Benchmarks & Datasets
CommonsenseQA
12,247 multiple choice questions requiring commonsense reasoning about everyday concepts.
N/A
Images
2019
Year
See Leaderboard
HellaSwag
70K sentence completion problems testing commonsense natural language inference.
N/A
Images
2019
Year
See Leaderboard
WinoGrande
44K Winograd-style problems requiring commonsense reasoning to resolve pronoun references.
N/A
Images
2019
Year
See Leaderboard
ARC-Challenge
7,787 science questions requiring reasoning. Challenge set contains harder questions that retrieval fails on.
N/A
Images
2018
Year
See Leaderboard
MMLU
15,908 multiple choice questions across 57 subjects from elementary to professional level.
N/A
Images
2021
Year
See Leaderboard