Research map / Large language models / Retrieval and knowledge
Embedding-Based Context-Aware Reranker
Ye Yuan, Mohammad Amin Shabani, Siqi Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add document ID embeddings and passage position encodings (ablation) | EBCAR · ConTEB Insurance | nDCG@10: 34.16 -> 40.74 | Improved |
| add document ID embeddings and passage position encodings (ablation) | EBCAR · ConTEB SQuAD | nDCG@10: 60.87 -> 71.62 | Improved |
| add document ID embeddings and passage position encodings (ablation) | EBCAR · ConTEB Football | nDCG@10: 42.88 -> 80.19 | Improved |
| add hybrid attention (shared full plus dedicated masked) (ablation) | EBCAR · ConTEB SQuAD | nDCG@10: 47.52 -> 71.62 | Improved |
| add hybrid attention (shared full plus dedicated masked) (ablation) | EBCAR · ConTEB Football | nDCG@10: 41.93 -> 80.19 | Improved |
| add hybrid attention (shared full plus dedicated masked) (ablation) | EBCAR · ConTEB MLDR | nDCG@10: 74.55 -> 75.26 | Improved |
| add both position info and hybrid attention (ablation) | EBCAR · ConTEB Football | nDCG@10: 5.28 -> 80.19 | Improved |
| add both position info and hybrid attention (ablation) | EBCAR · ConTEB Insurance | nDCG@10: 2.88 -> 40.74 | Improved |
| replace Contriever with stronger E5 embeddings | EBCAR · ConTEB Football | nDCG@10: 80.19 -> 80.64 | Improved |
| replace Contriever with stronger E5 embeddings | EBCAR · ConTEB MLDR | nDCG@10: 75.26 -> 77.36 | Improved |
| replace Contriever with stronger E5 embeddings | EBCAR · ConTEB ESG | nDCG@10: 37.20 -> 36.88 | No clear effect |
| use updated query embedding in contrastive loss instead of fixed query embedding (ablation) | EBCAR · training stability | nDCG@10 | Hurt |
Closest papers by meaning
- DynamicRAG: Leveraging Outputs of Large Language Model as Feedback for Dynamic Reranking in Retrieval-Augmented Generation (NeurIPS 2025 · similarity 0.89)
- Contrastive Retrieval Heads for Improved Attention-Based Reranking (NeurIPS 2026 · similarity 0.89)
- Attention in Large Language Models Yields Efficient Zero-Shot Re-Rankers (ICLR 2025 · similarity 0.89)
- AcuRank: Uncertainty-Aware Adaptive Computation for Listwise Reranking (NeurIPS 2025 · similarity 0.88)
- Scalable In-context Ranking with Generative Models (NeurIPS 2025 · similarity 0.87)
- xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
