@inproceedings{abdulsattar-ross-2026-arabic,
title = "{A}rabic Dialect Translation with Small {LLM}s: Enhancing through Reasoning-Oriented Reinforcement Learning",
author = "Abdulsattar, Sohaila and
Ross, Keith",
booktitle = "Proceedings of the 2nd Workshop on {NLP} for Languages Using {A}rabic Script",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.abjadnlp-1.11/",
pages = "84--99",
abstract = "Arabic dialect{\ensuremath{\leftrightarrow}}English machine translation remains difficult due to extreme dialect variation, inconsistent orthography, and limited parallel data. Moreover, dialect translation is often needed in remote regions or by economically-disadvantaged communities, which often operate in compute-constrained or offline settings. Motivated by these concerns, in this paper we explore optimizing Arabic dialect{\ensuremath{\leftrightarrow}}English translators that run over small LLMs, which could be implemented on small offline devices. We show that reasoning-oriented reinforcement learning can substantially improve small multilingual LLMs for Arabic dialect translation. Using the MADAR corpus, small Qwen-2.5 models trained with a think-then-translate template and optimized with Group-Relative Policy Optimization using a SacreBLEU reward outperform a much larger 7B baseline trained with supervised fine-tuning. The dialect-to-English BLEU score more than doubles from 17.4 to 34.9, while the English-to-dialect COMET score improves from 0.57 to 0.73."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="abdulsattar-ross-2026-arabic">
<titleInfo>
<title>Arabic Dialect Translation with Small LLMs: Enhancing through Reasoning-Oriented Reinforcement Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sohaila</namePart>
<namePart type="family">Abdulsattar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Keith</namePart>
<namePart type="family">Ross</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Arabic dialect\ensuremathłeftrightarrowEnglish machine translation remains difficult due to extreme dialect variation, inconsistent orthography, and limited parallel data. Moreover, dialect translation is often needed in remote regions or by economically-disadvantaged communities, which often operate in compute-constrained or offline settings. Motivated by these concerns, in this paper we explore optimizing Arabic dialect\ensuremathłeftrightarrowEnglish translators that run over small LLMs, which could be implemented on small offline devices. We show that reasoning-oriented reinforcement learning can substantially improve small multilingual LLMs for Arabic dialect translation. Using the MADAR corpus, small Qwen-2.5 models trained with a think-then-translate template and optimized with Group-Relative Policy Optimization using a SacreBLEU reward outperform a much larger 7B baseline trained with supervised fine-tuning. The dialect-to-English BLEU score more than doubles from 17.4 to 34.9, while the English-to-dialect COMET score improves from 0.57 to 0.73.</abstract>
<identifier type="citekey">abdulsattar-ross-2026-arabic</identifier>
<location>
<url>https://aclanthology.org/2026.abjadnlp-1.11/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>84</start>
<end>99</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Arabic Dialect Translation with Small LLMs: Enhancing through Reasoning-Oriented Reinforcement Learning
%A Abdulsattar, Sohaila
%A Ross, Keith
%S Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%F abdulsattar-ross-2026-arabic
%X Arabic dialect\ensuremathłeftrightarrowEnglish machine translation remains difficult due to extreme dialect variation, inconsistent orthography, and limited parallel data. Moreover, dialect translation is often needed in remote regions or by economically-disadvantaged communities, which often operate in compute-constrained or offline settings. Motivated by these concerns, in this paper we explore optimizing Arabic dialect\ensuremathłeftrightarrowEnglish translators that run over small LLMs, which could be implemented on small offline devices. We show that reasoning-oriented reinforcement learning can substantially improve small multilingual LLMs for Arabic dialect translation. Using the MADAR corpus, small Qwen-2.5 models trained with a think-then-translate template and optimized with Group-Relative Policy Optimization using a SacreBLEU reward outperform a much larger 7B baseline trained with supervised fine-tuning. The dialect-to-English BLEU score more than doubles from 17.4 to 34.9, while the English-to-dialect COMET score improves from 0.57 to 0.73.
%U https://aclanthology.org/2026.abjadnlp-1.11/
%P 84-99
Markdown (Informal)
[Arabic Dialect Translation with Small LLMs: Enhancing through Reasoning-Oriented Reinforcement Learning](https://aclanthology.org/2026.abjadnlp-1.11/) (Abdulsattar & Ross, AbjadNLP 2026)
ACL