Commonsense Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
80.61Client 1 Accuracy
6
May 28, 2026
83.78Accuracy
6
May 28, 2026
70.7Accuracy
6
May 26, 2026
76.1ARC-c
6
May 26, 2026
86.1Accuracy (n=15)
6
May 21, 2026
63.84HellaSWAG Normalized Accuracy
6
May 19, 2026
0.695IAA
6
May 12, 2026
82.14BoolQ Accuracy (Zero-shot)
6
Apr 23, 2026
32.2Clean Error
6
Apr 15, 2026
64.2Accuracy
6
Mar 12, 2026
51.11Accuracy
6
Mar 6, 2026
45.6Average Score
6
Mar 4, 2026
48.83Accuracy
6
Feb 26, 2026
49.28Accuracy
6
Feb 26, 2026
57.2Accuracy
6
Feb 26, 2026
54.36Score
6
Feb 26, 2026
90.06Accuracy
6
Feb 26, 2026
78Accuracy
6
Feb 26, 2026
65.4Accuracy
6
Feb 26, 2026
76.2Accuracy
6
Feb 26, 2026
66AUC
6
Feb 26, 2026
54.53Accuracy
6
Feb 26, 2026
57.4Accuracy
6
Feb 26, 2026
74.67Accuracy
6
Feb 26, 2026
72.96Accuracy
6
Feb 26, 2026