Sycophancy Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
91.9AUROC
16
Jun 8, 2026
98.3Accuracy
14
Jun 26, 2026
98.3Accuracy
12
Jun 26, 2026
98.3Accuracy
12
Jun 26, 2026
100Accuracy
12
Jun 26, 2026
0.732AUROC
12
Feb 26, 2026
10Sycophancy Rate
10
May 27, 2026
68Sycophancy Rate
10
May 27, 2026