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

IHEval Safety Defense

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hierarchical Instruction FollowingIHEval Safety Defense Aligned setting
Extract Accuracy (E)63.2
3
Hierarchical Instruction FollowingIHEval Safety Defense Reference setting
Extract Accuracy78.9
3
Showing 2 of 2 rows