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

PEARL

Benchmarks

Task NameDataset NameSOTA ResultTrend
Conversational RecommendationPEARL (test)
Recall@12.17
4
Single-Turn User SimulationPEARL (test)
BertScore0.9373
4
Multi-Turn User SimulationPEARL (test)
Success Rate (SR)95.31
4
Showing 3 of 3 rows