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

Standard non-adversarial dataset

Benchmarks

Task NameDataset NameSOTA ResultTrend
Safety Content DetectionStandard non-adversarial dataset e-commerce
TPR90
3
Safety Content DetectionStandard non-adversarial dataset Romance
TPR100
3
Safety Content DetectionStandard non-adversarial dataset Tax Authority
TPR100
3
Safety Content DetectionStandard non-adversarial dataset Racism
TPR99
3
Safety Content DetectionStandard non-adversarial dataset Elections
TPR99
3
Safety Content DetectionStandard non-adversarial dataset Anti-LGBTQ
TPR100
3
Safety Content DetectionStandard non-adversarial dataset Delusional
TPR90
3
Safety Content DetectionStandard non-adversarial dataset SQL Injection
TPR100
3
Safety Content DetectionStandard non-adversarial dataset Phishing
TPR98
3
Showing 9 of 9 rows