Toxicity Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
84.23Score
95
Jun 2, 2026
58Win Count
56
Feb 26, 2026
63.1Accuracy
52
Jun 2, 2026
1F1 Score
45
Jun 1, 2026
59.74Accuracy
42
Jun 2, 2026
80.6Toxicity Score
24
Feb 26, 2026
90.44AUC
21
Feb 26, 2026
0.57Oracle ECE
20
May 12, 2026
92.7Average Accuracy
19
Feb 26, 2026
91.66Accuracy
15
Apr 13, 2026
78.8WGA
14
Apr 15, 2026
91.3Accuracy
13
Feb 26, 2026
98.92Accuracy
12
Mar 4, 2026
87.56Accuracy
12
Mar 4, 2026
52.5F1 Score
11
Feb 26, 2026
80.8F1 Score
11
Feb 26, 2026
66.4F1 Score
11
Feb 26, 2026
60Implicit Score
10
Jun 2, 2026
96.05Accuracy
10
Apr 14, 2026
85.57Performance (Insert)
10
Mar 17, 2026
75.73Accuracy
10
Feb 26, 2026
87.06Accuracy
10
Feb 26, 2026
68.1F1 Score
10
Feb 26, 2026
91.1F1 Score
10
Feb 26, 2026
81.1F1 Score
10
Feb 26, 2026