@inproceedings{li-etal-2025-multilingual,
title = "Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness",
author = "Li, Bryan and
Luo, Fiona and
Haider, Samar and
Agashe, Adwait and
Li, Siyu and
Liu, Runqi and
Miao, Miranda Muqing and
Ramakrishnan, Shriya and
Yuan, Yuan and
Callison-Burch, Chris",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.219/",
doi = "10.18653/v1/2025.findings-acl.219",
pages = "4215--4241",
ISBN = "979-8-89176-256-5",
abstract = "The paradigm of retrieval-augmented generated (RAG) helps mitigate hallucinations of large language models (LLMs). However, RAG also introduces biases contained within the retrieved documents. These biases can be amplified in scenarios which are multilingual and culturally-sensitive, such as territorial disputes. We thus introduce BordIRLines, a dataset of territorial disputes paired with retrieved Wikipedia documents, across 49 languages. We evaluate the cross-lingual robustness of this RAG setting by formalizing several modes for multilingual retrieval. Our experiments on several LLMs show that incorporating perspectives from diverse languages can in fact improve robustness; retrieving multilingual documents best improves response consistency and decreases geopolitical bias over RAG with purely in-language documents. We also consider how RAG responses utilize presented documents, finding a much wider variance in the linguistic distribution of response citations, when querying in low-resource languages. Our further analyses investigate the various aspects of a cross-lingual RAG pipeline, from retrieval to document contents. We release our benchmark to support continued research towards equitable information access across languages, at https://huggingface.co/datasets/borderlines/bordirlines."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2025-multilingual">
<titleInfo>
<title>Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness</title>
</titleInfo>
<name type="personal">
<namePart type="given">Bryan</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fiona</namePart>
<namePart type="family">Luo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samar</namePart>
<namePart type="family">Haider</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Adwait</namePart>
<namePart type="family">Agashe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siyu</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Runqi</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Miranda</namePart>
<namePart type="given">Muqing</namePart>
<namePart type="family">Miao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shriya</namePart>
<namePart type="family">Ramakrishnan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuan</namePart>
<namePart type="family">Yuan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chris</namePart>
<namePart type="family">Callison-Burch</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>The paradigm of retrieval-augmented generated (RAG) helps mitigate hallucinations of large language models (LLMs). However, RAG also introduces biases contained within the retrieved documents. These biases can be amplified in scenarios which are multilingual and culturally-sensitive, such as territorial disputes. We thus introduce BordIRLines, a dataset of territorial disputes paired with retrieved Wikipedia documents, across 49 languages. We evaluate the cross-lingual robustness of this RAG setting by formalizing several modes for multilingual retrieval. Our experiments on several LLMs show that incorporating perspectives from diverse languages can in fact improve robustness; retrieving multilingual documents best improves response consistency and decreases geopolitical bias over RAG with purely in-language documents. We also consider how RAG responses utilize presented documents, finding a much wider variance in the linguistic distribution of response citations, when querying in low-resource languages. Our further analyses investigate the various aspects of a cross-lingual RAG pipeline, from retrieval to document contents. We release our benchmark to support continued research towards equitable information access across languages, at https://huggingface.co/datasets/borderlines/bordirlines.</abstract>
<identifier type="citekey">li-etal-2025-multilingual</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.219</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.219/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>4215</start>
<end>4241</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness
%A Li, Bryan
%A Luo, Fiona
%A Haider, Samar
%A Agashe, Adwait
%A Li, Siyu
%A Liu, Runqi
%A Miao, Miranda Muqing
%A Ramakrishnan, Shriya
%A Yuan, Yuan
%A Callison-Burch, Chris
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F li-etal-2025-multilingual
%X The paradigm of retrieval-augmented generated (RAG) helps mitigate hallucinations of large language models (LLMs). However, RAG also introduces biases contained within the retrieved documents. These biases can be amplified in scenarios which are multilingual and culturally-sensitive, such as territorial disputes. We thus introduce BordIRLines, a dataset of territorial disputes paired with retrieved Wikipedia documents, across 49 languages. We evaluate the cross-lingual robustness of this RAG setting by formalizing several modes for multilingual retrieval. Our experiments on several LLMs show that incorporating perspectives from diverse languages can in fact improve robustness; retrieving multilingual documents best improves response consistency and decreases geopolitical bias over RAG with purely in-language documents. We also consider how RAG responses utilize presented documents, finding a much wider variance in the linguistic distribution of response citations, when querying in low-resource languages. Our further analyses investigate the various aspects of a cross-lingual RAG pipeline, from retrieval to document contents. We release our benchmark to support continued research towards equitable information access across languages, at https://huggingface.co/datasets/borderlines/bordirlines.
%R 10.18653/v1/2025.findings-acl.219
%U https://aclanthology.org/2025.findings-acl.219/
%U https://doi.org/10.18653/v1/2025.findings-acl.219
%P 4215-4241
Markdown (Informal)
[Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness](https://aclanthology.org/2025.findings-acl.219/) (Li et al., Findings 2025)
ACL
- Bryan Li, Fiona Luo, Samar Haider, Adwait Agashe, Siyu Li, Runqi Liu, Miranda Muqing Miao, Shriya Ramakrishnan, Yuan Yuan, and Chris Callison-Burch. 2025. Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness. In Findings of the Association for Computational Linguistics: ACL 2025, pages 4215–4241, Vienna, Austria. Association for Computational Linguistics.