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

Summarize from Human Feedback

Benchmarks

Task NameDataset NameSOTA ResultTrend
Personalized Preference ModelingSummarize from Human Feedback (test)
Mean Accuracy66.01
12
Personalized SummarizationSummarize from Human Feedback (test)
Win Rate78.23
9
Showing 2 of 2 rows