@inproceedings{lu-etal-2026-empirical,
title = "An Empirical Study of Many-Shot In-Context Learning for Machine Translation of Low-Resource Languages",
author = "Lu, Yinhan and
Jhajj, Gaganpreet and
Zhang, Chen and
Andy, Anietie and
Adelani, David Ifeoluwa",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {IJCNLP}-{AACL} 2026",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-aacl.142/",
pages = "2772--2804",
ISBN = "979-8-89176-444-6",
abstract = "In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks from a few examples, making it promising for languages underrepresented in pre-training. Recent work on many-shot ICL suggests that modern LLMs can further benefit from more ICL examples enabled by their long context windows. However, the inference cost can be prohibitive for low-resource language communities, and it is unclear whether example selection can reduce the number of examples needed. In this paper, we present a systematic empirical study of many-shot ICL for machine translation from English into ten truly low-resource languages recently added to FLORES+. We analyze the effects of retrieving more informative examples, using out-of-domain data, and ordering examples. Our findings show that many-shot ICL becomes more effective as the number of examples increases. More importantly, we show that BM25 retrieval substantially improves data efficiency: 50 retrieved examples roughly match 250 many-shot examples, while 250 retrieved examples approach the performance of 1,000 many-shot examples. We further show that example ordering has little effect and that ICL provides additional gains on top of fine-tuning."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lu-etal-2026-empirical">
<titleInfo>
<title>An Empirical Study of Many-Shot In-Context Learning for Machine Translation of Low-Resource Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yinhan</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gaganpreet</namePart>
<namePart type="family">Jhajj</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chen</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anietie</namePart>
<namePart type="family">Andy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="given">Ifeoluwa</namePart>
<namePart type="family">Adelani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-444-6</identifier>
</relatedItem>
<abstract>In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks from a few examples, making it promising for languages underrepresented in pre-training. Recent work on many-shot ICL suggests that modern LLMs can further benefit from more ICL examples enabled by their long context windows. However, the inference cost can be prohibitive for low-resource language communities, and it is unclear whether example selection can reduce the number of examples needed. In this paper, we present a systematic empirical study of many-shot ICL for machine translation from English into ten truly low-resource languages recently added to FLORES+. We analyze the effects of retrieving more informative examples, using out-of-domain data, and ordering examples. Our findings show that many-shot ICL becomes more effective as the number of examples increases. More importantly, we show that BM25 retrieval substantially improves data efficiency: 50 retrieved examples roughly match 250 many-shot examples, while 250 retrieved examples approach the performance of 1,000 many-shot examples. We further show that example ordering has little effect and that ICL provides additional gains on top of fine-tuning.</abstract>
<identifier type="citekey">lu-etal-2026-empirical</identifier>
<location>
<url>https://aclanthology.org/2026.findings-aacl.142/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>2772</start>
<end>2804</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T An Empirical Study of Many-Shot In-Context Learning for Machine Translation of Low-Resource Languages
%A Lu, Yinhan
%A Jhajj, Gaganpreet
%A Zhang, Chen
%A Andy, Anietie
%A Adelani, David Ifeoluwa
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F lu-etal-2026-empirical
%X In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks from a few examples, making it promising for languages underrepresented in pre-training. Recent work on many-shot ICL suggests that modern LLMs can further benefit from more ICL examples enabled by their long context windows. However, the inference cost can be prohibitive for low-resource language communities, and it is unclear whether example selection can reduce the number of examples needed. In this paper, we present a systematic empirical study of many-shot ICL for machine translation from English into ten truly low-resource languages recently added to FLORES+. We analyze the effects of retrieving more informative examples, using out-of-domain data, and ordering examples. Our findings show that many-shot ICL becomes more effective as the number of examples increases. More importantly, we show that BM25 retrieval substantially improves data efficiency: 50 retrieved examples roughly match 250 many-shot examples, while 250 retrieved examples approach the performance of 1,000 many-shot examples. We further show that example ordering has little effect and that ICL provides additional gains on top of fine-tuning.
%U https://aclanthology.org/2026.findings-aacl.142/
%P 2772-2804
Markdown (Informal)
[An Empirical Study of Many-Shot In-Context Learning for Machine Translation of Low-Resource Languages](https://aclanthology.org/2026.findings-aacl.142/) (Lu et al., Findings 2026)
ACL