Common Sense Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
99.2Accuracy
288
Jun 29, 2026
92.4Accuracy
280
Jul 7, 2026
93.87Accuracy
213
Apr 7, 2026
91.3Accuracy
189
Mar 30, 2026
93.27ARC (easy) Accuracy
149
Jul 7, 2026
91.89Accuracy
100
May 7, 2026
73.410-shot Accuracy
92
Jun 8, 2026
77.67Average Accuracy (Zero-Shot)
92
Mar 27, 2026
83.9Accuracy
86
Jun 24, 2026
91.84Accuracy
85
Jul 9, 2026
69.46Accuracy
74
Jul 7, 2026
85.64Accuracy
67
Jul 7, 2026
89.6Accuracy
63
Jul 7, 2026
71.72Average Performance
61
Feb 26, 2026
94.1Accuracy
61
Jun 26, 2026
0.975Accuracy
52
Feb 26, 2026
84.4Accuracy
49
Jun 26, 2026
95.7Accuracy (acc_n)
47
Jun 2, 2026
90.4Average Accuracy
47
Jul 2, 2026
79.5Accuracy
43
Jun 26, 2026
60Accuracy
40
Feb 26, 2026
87.07Average Accuracy
35
Jul 2, 2026
68.95Accuracy
32
Jun 2, 2026
57.2Accuracy
31
Apr 9, 2026
79.35PIQA
30
Feb 26, 2026