Zero-shot Commonsense Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.87ARC-e Accuracy
43
May 13, 2026
73.18BoolQ Accuracy
32
Mar 17, 2026
76.93Accuracy
28
Jun 26, 2026
59Zero-shot Accuracy
21
May 18, 2026
61Average Accuracy
12
Feb 26, 2026
45.5ARC-E Accuracy
9
Mar 20, 2026
44.7HellaSwag Accuracy
7
Mar 25, 2026
83.19PIQA
6
Feb 26, 2026
67.17Accuracy (zero-shot)
4
Apr 23, 2026