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

CRSArena-Eval

Benchmarks

Task NameDataset NameSOTA ResultTrend
Reference-free Conversation EvaluationCRSArena-Eval KG
Pearson Correlation (r)0.766
44
Reference-free Conversation EvaluationCRSArena-Eval RD
Pearson Correlation (r)0.712
44
Dialogue Evaluation AlignmentCRSArena-Eval KG (test)
Pearson r (Turn-level Rel.)0.543
6
Dialogue Evaluation AlignmentCRSArena-Eval RD (test)
Pearson r (Turn-level Rel)0.549
6
Showing 4 of 4 rows