Misalignment Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
99Accuracy
63
May 28, 2026
52Average IR
30
Jul 3, 2026
31.4Avg. IR
30
Jul 3, 2026
0.3Error Rate (ER)
30
Jul 3, 2026
0.8Error Rate (ER)
30
Jul 3, 2026
89.4Accuracy
6
Apr 28, 2026
97.2Accuracy
6
Apr 28, 2026
93AUROC
4
Jun 24, 2026
93.5AUROC
4
Jun 24, 2026
0.817AUROC
4
Jun 24, 2026
95.6AUROC
4
Jun 24, 2026
0.985AUROC
4
Jun 24, 2026
0.959AUROC
4
Jun 24, 2026