Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

RecIF-Bench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Label-Conditional RecommendationRecIF-Bench Label-Cond. Rec
Pass@320.0574
20
Product RecommendationRecIF-Bench Product Rec
Pass@12.31
20
Ad RecommendationRecIF-Bench Ad Rec
Pass@10.0273
20
Short Video RecommendationRecIF-Bench Short Video Rec
Pass@15.74
20
Interactive RecommendationRecIF-Bench Interactive Rec
Pass@113.1
11
Label PredictionRecIF-Bench Label Pred
AUC0.6912
11
Label PredictionRecIF-Bench
AUC0.7017
9
Recommendation ExplanationRecIF-Bench Rec. Explanation
LLM Judge Score4.0381
5
Item UnderstandingRecIF-Bench Item Understand
LLM Judge Score0.3209
5
RecommendationRecIF-Bench Short Video
Entropy@104.455
3
Showing 10 of 10 rows