Logical reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
68.9Accuracy (LogiQA)
12
May 27, 2026
96Accuracy (Level 3)
12
May 14, 2026
49Accuracy
12
Mar 24, 2026
0.34Accuracy
12
Mar 24, 2026
60.3Accuracy
12
Feb 26, 2026
100Track(Avg)
12
Feb 27, 2026
24.5Accuracy
12
Feb 27, 2026
100Activation Score
11
Jun 30, 2026
66.5Logical Deduction Accuracy
11
May 27, 2026
77.5Accuracy
11
Feb 26, 2026
74.1Accuracy
11
Feb 26, 2026
86Accuracy
11
Feb 26, 2026
36.5Pass@1 Accuracy
10
Jun 25, 2026
1Normalized Generalization Error Approximation
10
Jun 15, 2026
1.1Generalization Error Approx. (G(W))
10
Jun 15, 2026
69Pass@1 Success Rate
10
May 25, 2026
98Pass@1
10
May 25, 2026
61.8Pass@1
10
Apr 14, 2026
52.4Pass@1 Accuracy
10
Apr 14, 2026
53Pass@1 Accuracy
10
Apr 14, 2026
95.7Accuracy
10
Mar 24, 2026
0.761Average PS
9
Jun 4, 2026
48.75Task Success Rate (TSR)
9
May 28, 2026
0.038Mean Improvement
9
May 18, 2026
25Jigsaw Score
9
Feb 26, 2026