ML
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Click-Through Rate Prediction | ML 1M | AUC0.9087 | 46 | |
| Recommendation | Ml 1M (test) | Recall21.2 | 24 | |
| Rubric satisfaction evaluation | ML | Claude-4 Sonnet Score36.7 | 24 | |
| Negative Constraint Recommendation | ML 1M | Recall@100.2076 | 22 | |
| Recommendation | ML-100K | HR@546.15 | 18 | |
| Sequential Recommendation | ML-20M | Memory (MB)2,315 | 18 | |
| Collaborative Filtering | ML-20M large (test) | Recall@200.403 | 17 | |
| Collaborative Filtering | ML-20M strong generalization | AOA Recall@200.3956 | 14 | |
| Model Extraction | ML-1M (test) | N@1062.6 | 12 | |
| Recommendation | ML-100K | NDCG@119.21 | 11 | |
| CTR Prediction | ML-1M | AUC0.8194 | 11 | |
| Recommendation | ML-10M (test) | RMSE0.777 | 10 | |
| Federated Recommendation and Attribute Unlearning | ML-100K Age attribute 1.0 (leave-one-out) | HR@1067.03 | 9 | |
| User behavior simulation | ML-100K | Precision71.99 | 9 | |
| Collaborative Filtering | ML-1M | HR@1031.79 | 9 | |
| Recommendation System Efficiency | ML 1M (overall) | Training Time (m)2.06 | 9 | |
| Item Response Theory Assessment | ML-1M | AUC0.701 | 9 | |
| Generative Recommendation | ML 20M | NDCG@100.1233 | 8 | |
| Positive Constraint Recommendation | ML1M | Recall@1073 | 8 | |
| Collaborative Filtering | ML-10M | HR@100.3676 | 8 | |
| Recommendation System Efficiency | ML 10M (overall) | Training Time (h)3.06 | 8 | |
| Sequential Recommendation | ML-1M implicit feedback | HR@519.44 | 8 | |
| Recommendation | ML-20M (test) | R@1096.2 | 8 | |
| Future item recommendation | ML-100K | Recall14.9 | 8 | |
| Ranking | ML-20M (test) | Recall@2039.5 | 8 |
Showing 25 of 44 rows