XProvence: Zero-Cost Multilingual Context Pruning for Retrieval-Augmented Generation
This work addresses the need for efficient multilingual RAG systems, though it is incremental as it extends an existing English framework to multiple languages.
The paper tackles the problem of multilingual context pruning for retrieval-augmented generation (RAG) by introducing XProvence, a model trained on 16 languages and supporting 100+ languages, which prunes RAG contexts with minimal-to-no performance degradation and outperforms strong baselines on four multilingual question answering benchmarks.
This paper introduces XProvence, a multilingual zero-cost context pruning model for retrieval-augmented generation (RAG), trained on 16 languages and supporting 100+ languages through effective cross-lingual transfer. Motivated by the growing use of RAG systems across diverse languages, we explore several strategies to generalize the Provence framework-which first integrated efficient zero-cost context pruning directly into the re-ranking model-beyond English. Across four multilingual question answering benchmarks, we show how XProvence can prune RAG contexts with minimal-to-no performance degradation and outperforms strong baselines. Our model is available at https://huggingface.co/naver/xprovence-reranker-bgem3-v2.