Commonsense Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
76.84Accuracy
20
May 26, 2026
88.1Accuracy
20
May 26, 2026
81.19ARC-E Accuracy
20
May 18, 2026
48.89LM Eval Score
20
Apr 21, 2026
58ARC-C Accuracy (5-shot)
20
Feb 26, 2026
64.2Accuracy
19
May 26, 2026
89.3Accuracy
19
Jun 19, 2026
0.862Accuracy
19
Feb 26, 2026
73.2Accuracy (Pre-Attack)
19
Feb 26, 2026
66.8Accuracy
18
Jun 16, 2026
66.92PIQA Accuracy
18
May 12, 2026
66.4Accuracy (CSQA)
18
Mar 31, 2026
57.8Accuracy
18
Jul 3, 2026
43.73Accuracy
18
Feb 26, 2026
79.68Average Accuracy
18
Feb 26, 2026
76.45Accuracy
18
Feb 26, 2026
71.7Accuracy
18
Feb 26, 2026
78PIQA Accuracy
17
May 18, 2026
82.9Accuracy
17
May 8, 2026
89.1Accuracy
17
May 8, 2026
84.98Accuracy
17
Apr 6, 2026
80.92BoolQ Accuracy (Zero-shot)
17
Feb 26, 2026
91.88Accuracy
17
Feb 26, 2026
75.1Accuracy
17
Feb 26, 2026
85.6Accuracy
17
Feb 26, 2026