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

HH and UF

Benchmarks

Task NameDataset NameSOTA ResultTrend
Preference AlignmentHH and UF Out-of-Domain (test)
OOD Win Rate58.7
10
Preference AlignmentHH and UF In-Domain (test)
Win Rate68.4
10
Showing 2 of 2 rows