Commonsense Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
60.66Accuracy
16
Apr 23, 2026
63.27BoolQ Accuracy
16
Apr 13, 2026
88.8Accuracy
16
Mar 4, 2026
83.1Accuracy
16
Feb 26, 2026
80.4Accuracy
16
Feb 26, 2026
85.42Accuracy
16
Feb 26, 2026
93.9ARC-c Accuracy
15
Apr 24, 2026
38.29HellaSwag (Zero-shot)
15
Apr 1, 2026
80.79PiQA Accuracy
15
Feb 26, 2026
82.21PiQA Accuracy
15
Feb 26, 2026
82.35Accuracy
15
Feb 26, 2026
81.8Accuracy
15
Feb 26, 2026
75.6Accuracy (Inter-layer)
15
Feb 26, 2026
67Accuracy (Inter-Layer Filtering)
15
Feb 26, 2026
69.5Accuracy (StrategyQA)
14
Jun 19, 2026
44.2OBQA Accuracy
14
Jun 9, 2026
80.6Pass Rate
14
May 26, 2026
82.6Accuracy
14
May 8, 2026
76.19Accuracy
14
Apr 28, 2026
87.27Average Accuracy
14
Apr 28, 2026
76.53Accuracy
14
Mar 6, 2026
67.3CSR Accuracy
14
Jun 5, 2026
79.3Accuracy
14
Feb 26, 2026
84.3Accuracy
14
Feb 26, 2026
78.05Vanilla Acc
14
Feb 26, 2026