Amazon
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Amazon Hard
45Kimi K2 Success Rate
5
Amazon Easy
87.5Kimi K2 Success Rate
5
Amazon Health (test)
3.25Recall@20
5
Amazon Electronics & Phones (test)
688Avg Local Train Time (s)
5
Amazon Phones (test)
2.16H@5
5
Amazon Electronics (test)
1.3H@5
5
Amazon-Book
0.731AUC
5
Amazon Toys
11.2HR@50
5
Amazon Sports
7.2HR@50
5
Amazon 5-core Movies and TV
0.661MAE
5
Amazon 5-core CDs and Vinyl
0.575MAE
5
Amazon 5-core Health and Personal Care
0.642MAE
5
Amazon 5-core Video Games
0.715MAE
5
Amazon Digital Music 5-core
0.563MAE
5
AMAZON-13K
93.672nDCG
5
Amazon
68.34CLIP Score
5
Amazon-all large-scale (test)
84.76AUROC
5
Amazon Book
0.159Recall@20
5
Amazon -> SST-2 (target set)
0.929Target F1
5
Amazon (test)
67.24Accuracy
5
Amazon
0.95RMSE
5
AMAZON preprocessed (test)
0.3Runtime (min)
5
Amazon-12K
94.87Precision@1
5
Amazon SNAP collection (test)
97Average Accuracy
5
Amazon Beauty (test)
23.9Precision
4