All Languages Matter: Understanding and Mitigating Language Bias in Multilingual RAG
On the map: Retrieval-augmented generation → Generation and context integration
Abstract (arXiv)
Multilingual Retrieval-Augmented Generation (mRAG) leverages cross-lingual evidence to ground Large Language Models (LLMs) in global knowledge. However, we show that current mRAG systems suffer from a language bias during reranking, systematically favoring English and the query's native language. By introducing an estimated oracle evidence analysis, we quantify a substantial performance gap between existing rerankers and the achievable upper bound. Further analysis reveals a critical distributional mismatch: while optimal predictions require evidence scattered across multiple languages, current systems systematically suppress such ``answer-critical'' documents, thereby limiting downstream generation performance. To bridge this gap, we propose textitLanguage-Agnostic Utility-driven Reranker Alignment (LAURA), which aligns multilingual evidence ranking with downstream generative utility. Experiments across diverse languages and generation models show that LAURA effectively mitigates language bias and consistently improves mRAG performance.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Generation and context integration, one of 15 directions in Retrieval-augmented generation. Builds on 16 of its 23 references that are in the map. First posted on arXiv 22 Apr 2026.
Builds on
16 of 23 references.
Closest ancestors · 7
Same direction, newest first
- 2025Quality-Aware Translation Tagging in Multilingual RAG system27 Oct
- 2025The Cross-Lingual Cost: Retrieval Biases in RAG over Arabic-English Corpora10 Jul
- 2025On the Consistency of Multilingual Context Utilization in Retrieval-Augmented Generation1 Apr
- 2025Investigating Language Preference of Multilingual RAG Systems16 Feb
- 2023In-Context Retrieval-Augmented Language Models31 Jan
- 2022Retrieval-Augmented Multilingual Keyphrase Generation with Retriever-Generator Iterative Training21 May
- 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks22 May
Other references · 9
- 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,175
- 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,050
- 2024SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuningcited by 345
- 2020MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answeringcited by 151
- 2020XOR QA: Cross-lingual Open-Retrieval Question Answeringcited by 121
- 2021One Question Answering Model for Many Languages with Cross-lingual Dense Passage Retrievalcited by 46
- 2024Retrieval-augmented generation in multilingual settingscited by 27
- 2022Towards Best Practices for Training Multilingual Dense Retrieval Modelscited by 26
- 2024Language Fairness in Multilingual Information Retrievalcited by 5
Built on by
Citing papers per year: 2026: 2.
2 papers.
Similar, not cited
Close in wording, no citation either way
- 2026Enhancing Multilingual RAG Systems with Debiased Language Preference-Guided Query Fusion
- 2025Language Drift in Multilingual Retrieval-Augmented Generation: Characterization and Decoding-Time Mitigation
- 2024Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness
- 2023"Knowing When You Don't Know": A Multilingual Relevance Assessment Dataset for Robust Retrieval-Augmented Generation
- 2025Improving Multilingual Retrieval-Augmented Language Models through Dialectic Reasoning Argumentations
- 2025Multilingual Retrieval-Augmented Generation for Knowledge-Intensive Task
- 2024MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries
- 2025XRAG: Cross-lingual Retrieval-Augmented Generation