Toxicity Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
22F1 Score
10
Feb 26, 2026
41.3F1 Score
10
Feb 26, 2026
0Accuracy Gap (AG)
8
Apr 13, 2026
5.4EOD
8
Apr 13, 2026
78.72Accuracy
8
Feb 26, 2026
73.42Accuracy
8
Feb 26, 2026
96AUC
8
Feb 26, 2026
0.66AUC
8
Feb 26, 2026
0.67AUC
8
Feb 26, 2026
0.71AUC
8
Feb 26, 2026
91.78ASR
7
May 22, 2026
90.33ASR
7
May 22, 2026
84.14ASR
7
May 22, 2026
85.2ASR
7
May 22, 2026
88.7ASR
7
May 22, 2026
90.2ASR
7
May 22, 2026
80.4ASR
7
May 22, 2026
75.3ASR
7
May 22, 2026
87.7Accuracy (LGBTQ+)
7
Feb 26, 2026
87Accuracy
7
Feb 26, 2026
84Accuracy
7
Feb 26, 2026
97.6AUROC
6
May 13, 2026
73Macro-F1
6
Apr 8, 2026
92.3Average Accuracy
5
Feb 26, 2026
98.86ROC Score
4
Jun 26, 2026