# All Languages Matter: Understanding and Mitigating Language Bias in Multilingual RAG

arXiv: https://arxiv.org/abs/2604.20199 · HTML: https://mlsanity.com/paper/2604.20199/

## Key facts

- Cited by 3 papers in the map (data through 14 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.

## Placed in

- Retrieval-augmented generation / [Generation and context integration](https://mlsanity.com/rag/generation-and-context-integration/)

## Builds on (most cited first, 16 of 16)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation](https://mlsanity.com/paper/2402.03216/) (2024)
- [In-Context Retrieval-Augmented Language Models](https://mlsanity.com/paper/2302.00083/) (2023)
- [SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuning](https://mlsanity.com/paper/2408.05517/) (2024)
- [MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering](https://mlsanity.com/paper/2007.15207/) (2020)
- [XOR QA: Cross-lingual Open-Retrieval Question Answering](https://mlsanity.com/paper/2010.11856/) (2020)
- [One Question Answering Model for Many Languages with Cross-lingual Dense Passage Retrieval](https://mlsanity.com/paper/2107.11976/) (2021)
- [Retrieval-augmented generation in multilingual settings](https://mlsanity.com/paper/2407.01463/) (2024)
- [Towards Best Practices for Training Multilingual Dense Retrieval Models](https://mlsanity.com/paper/2204.02363/) (2022)
- [Investigating Language Preference of Multilingual RAG Systems](https://mlsanity.com/paper/2502.11175/) (2025)
- [Retrieval-Augmented Multilingual Keyphrase Generation with Retriever-Generator Iterative Training](https://mlsanity.com/paper/2205.10471/) (2022)
- [On the Consistency of Multilingual Context Utilization in Retrieval-Augmented Generation](https://mlsanity.com/paper/2504.00597/) (2025)
- [Language Fairness in Multilingual Information Retrieval](https://mlsanity.com/paper/2405.00978/) (2024)
- [The Cross-Lingual Cost: Retrieval Biases in RAG over Arabic-English Corpora](https://mlsanity.com/paper/2507.07543/) (2025)
- [Quality-Aware Translation Tagging in Multilingual RAG system](https://mlsanity.com/paper/2510.23070/) (2025)

## Built on by (most cited first, 3 of 3)

- [LAMAR: An Open Language-Aware Multilingual Alignment Reranker](https://mlsanity.com/paper/2607.22042/) (2026)
- [Where's Waldo? Query-language Preference under Cross-lingual Knowledge Disparities](https://mlsanity.com/paper/2610.00606/) (2026)
- [Retrieval Disagreement as a Signal: Diagnosing and Steering Multilingual Dense Embedders](https://mlsanity.com/paper/2610.18526/) (2026)

## 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.