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

EDU

Benchmarks

Task NameDataset NameSOTA ResultTrend
Preference AlignmentEDU
Human Win Rate39.1
8
Worst-slab coverage (WSC)EDU
WSC Coverage80
6
Node RegressionEDU (test)
Marginal Coverage93.7
6
Showing 3 of 3 rows