Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

COLD

Benchmarks

Task NameDataset NameSOTA ResultTrend
Toxicity ClassificationCOLD (test)
Accuracy94.33
19
Content SafetyCOLD (test)
Macro F1-Score74.8
8
Toxicity DetectionCOLD
ASR75.3
7
Showing 3 of 3 rows