@inproceedings{zebaze-etal-2025-context,
title = "In-Context Example Selection via Similarity Search Improves Low-Resource Machine Translation",
author = "Zebaze, Armel Randy and
Sagot, Beno{\^i}t and
Bawden, Rachel",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.68/",
doi = "10.18653/v1/2025.findings-naacl.68",
pages = "1222--1252",
ISBN = "979-8-89176-195-7",
abstract = "The ability of generative large language models (LLMs) to perform in-context learning has given rise to a large body of research into how best to prompt models for various natural language processing tasks. In this paper, we focus on machine translation (MT), a task that has been shown to benefit from in-context translation examples. However no systematic studies have been published on how best to select examples, and mixed results have been reported on the usefulness of similarity-based selection over random selection, although these results have mainly been shown for high-resource languages only. We provide a study covering multiple LLMs and in-context example retrieval strategies. Contrarily to previously published results, we find that retrieval based on sentence embedding similarity can improve MT, especially for low-resource language directions, and we also discuss the balance between selection pool diversity and quality. Code and outputs will be made freely available."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zebaze-etal-2025-context">
<titleInfo>
<title>In-Context Example Selection via Similarity Search Improves Low-Resource Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Armel</namePart>
<namePart type="given">Randy</namePart>
<namePart type="family">Zebaze</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Benoît</namePart>
<namePart type="family">Sagot</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rachel</namePart>
<namePart type="family">Bawden</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>The ability of generative large language models (LLMs) to perform in-context learning has given rise to a large body of research into how best to prompt models for various natural language processing tasks. In this paper, we focus on machine translation (MT), a task that has been shown to benefit from in-context translation examples. However no systematic studies have been published on how best to select examples, and mixed results have been reported on the usefulness of similarity-based selection over random selection, although these results have mainly been shown for high-resource languages only. We provide a study covering multiple LLMs and in-context example retrieval strategies. Contrarily to previously published results, we find that retrieval based on sentence embedding similarity can improve MT, especially for low-resource language directions, and we also discuss the balance between selection pool diversity and quality. Code and outputs will be made freely available.</abstract>
<identifier type="citekey">zebaze-etal-2025-context</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.68</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.68/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>1222</start>
<end>1252</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T In-Context Example Selection via Similarity Search Improves Low-Resource Machine Translation
%A Zebaze, Armel Randy
%A Sagot, Benoît
%A Bawden, Rachel
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F zebaze-etal-2025-context
%X The ability of generative large language models (LLMs) to perform in-context learning has given rise to a large body of research into how best to prompt models for various natural language processing tasks. In this paper, we focus on machine translation (MT), a task that has been shown to benefit from in-context translation examples. However no systematic studies have been published on how best to select examples, and mixed results have been reported on the usefulness of similarity-based selection over random selection, although these results have mainly been shown for high-resource languages only. We provide a study covering multiple LLMs and in-context example retrieval strategies. Contrarily to previously published results, we find that retrieval based on sentence embedding similarity can improve MT, especially for low-resource language directions, and we also discuss the balance between selection pool diversity and quality. Code and outputs will be made freely available.
%R 10.18653/v1/2025.findings-naacl.68
%U https://aclanthology.org/2025.findings-naacl.68/
%U https://doi.org/10.18653/v1/2025.findings-naacl.68
%P 1222-1252
Markdown (Informal)
[In-Context Example Selection via Similarity Search Improves Low-Resource Machine Translation](https://aclanthology.org/2025.findings-naacl.68/) (Zebaze et al., Findings 2025)
ACL