Logical reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
100Accuracy (All)
4
Jun 16, 2026
100Boolean Expressions Accuracy
4
May 27, 2026
23.88Accuracy (128 tokens)
4
May 15, 2026
4.2Avg@4 Score
4
May 12, 2026
71.2Accuracy
4
Apr 14, 2026
68.1Accuracy
4
Apr 9, 2026
98.01Accuracy
4
Mar 6, 2026
72.734Accuracy
4
Feb 26, 2026
75.6Accuracy
4
Feb 26, 2026
49.6Accuracy
4
Feb 26, 2026
80.8Accuracy
4
Mar 10, 2026
0.722AUROC
3
Jul 9, 2026
18.37Accuracy
3
May 19, 2026
41.6Accuracy
3
May 12, 2026
34.5Accuracy
3
May 12, 2026
45.2Accuracy
3
May 12, 2026
35.6Accuracy
3
May 12, 2026
25Accuracy
3
May 12, 2026
24Accuracy
3
May 12, 2026
0.72Per-Class Loss
3
May 8, 2026
44.8Accuracy
3
May 8, 2026
72.4Accuracy
3
May 8, 2026
0Error Rate
3
Feb 26, 2026
100Accuracy
3
Feb 26, 2026
453Optimization-phase Token Usage
3
Feb 26, 2026