ESCI
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Adversarial Attack Effectiveness in RAG Pipeline | ESCI position 6 (test) | Success Rate @10100 | 28 | |
| GEO Attack Detection | ESCI Pipeline (≈1:9) (test) | FDR0 | 28 | |
| GEO Attack Detection | ESCI Balanced (1:1) (test) | FDR0 | 28 | |
| Retrieval | Amazon ESCI us (test) | Recall@57.22 | 14 | |
| Search Relevance | ESCI | Macro F161.03 | 14 | |
| Retrieval | Amazon ESCI jp (test) | Recall@55.57 | 13 | |
| Retrieval | Amazon ESCI es (test) | R@56.06 | 13 | |
| Relevance Classification | ESCI-ES (test) | ACC68.1 | 13 | |
| Relevance Classification | ESCI-US (test) | Accuracy70.23 | 13 | |
| Product Search | ESCI | NDCG (Two-C)82.59 | 9 | |
| Product Search | ESCI | NDCG@1058.8 | 8 | |
| Query Rewriting | ESCI | Coverage (cov(Q))171 | 6 | |
| Precision@K estimation | ESCI | Bias0.29 | 3 | |
| Product Search | ESCI (test) | NDCG (Two-C)- | 0 |
Showing 14 of 14 rows