CoRAG: Collaborative Retrieval-Augmented Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912305952849920 |
|---|---|
| author | Muhamed, Aashiq Diab, Mona Smith, Virginia |
| author_facet | Muhamed, Aashiq Diab, Mona Smith, Virginia |
| contents | Retrieval-Augmented Generation (RAG) models excel in knowledge-intensive tasks, especially under few-shot learning constraints. We introduce CoRAG, a framework extending RAG to collaborative settings, where clients jointly train a shared model using a collaborative passage store. To evaluate CoRAG, we introduce CRAB, a benchmark for collaborative homogeneous open-domain question answering. Our experiments demonstrate that CoRAG consistently outperforms both parametric collaborative learning methods and locally trained RAG models in low-resource scenarios. Further analysis reveals the critical importance of relevant passages within the shared store, the surprising benefits of incorporating irrelevant passages, and the potential for hard negatives to negatively impact performance. This introduces a novel consideration in collaborative RAG: the trade-off between leveraging a collectively enriched knowledge base and the potential risk of incorporating detrimental passages from other clients. Our findings underscore the viability of CoRAG, while also highlighting key design challenges and promising avenues for future research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_01883 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CoRAG: Collaborative Retrieval-Augmented Generation Muhamed, Aashiq Diab, Mona Smith, Virginia Artificial Intelligence Computation and Language Information Retrieval Machine Learning Retrieval-Augmented Generation (RAG) models excel in knowledge-intensive tasks, especially under few-shot learning constraints. We introduce CoRAG, a framework extending RAG to collaborative settings, where clients jointly train a shared model using a collaborative passage store. To evaluate CoRAG, we introduce CRAB, a benchmark for collaborative homogeneous open-domain question answering. Our experiments demonstrate that CoRAG consistently outperforms both parametric collaborative learning methods and locally trained RAG models in low-resource scenarios. Further analysis reveals the critical importance of relevant passages within the shared store, the surprising benefits of incorporating irrelevant passages, and the potential for hard negatives to negatively impact performance. This introduces a novel consideration in collaborative RAG: the trade-off between leveraging a collectively enriched knowledge base and the potential risk of incorporating detrimental passages from other clients. Our findings underscore the viability of CoRAG, while also highlighting key design challenges and promising avenues for future research. |
| title | CoRAG: Collaborative Retrieval-Augmented Generation |
| topic | Artificial Intelligence Computation and Language Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2504.01883 |