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

Harm Categories

Benchmarks

Task NameDataset NameSOTA ResultTrend
Safety EvaluationHarm Categories evaluated by Claude-Opus-4-7 (Average)
Safety Score95.3
6
Pairwise RankingSeven Harm Categories
Insult Pairwise Score83.1
3
Showing 2 of 2 rows