@inproceedings{al-kautsar-koto-2026-parallel,
title = "Parallel Tokenizers: Rethinking Vocabulary Design in Cross-Lingual Transfer of Low-Resource Languages",
author = "Al Kautsar, Muhammad Dehan and
Koto, Fajri",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.117/",
pages = "2718--2734",
ISBN = "979-8-89176-459-0",
abstract = "Tokenization forms the basis of multilingual language models, yet existing methods often limit cross-lingual transfer by mapping semantically equivalent words to different embeddings. For example, `I eat rice' in English and `Ina cin shinkafa' in Hausa are typically mapped to different vocabulary indices, preventing shared representations and limiting cross-lingual generalization. This problem is even more pronounced in low-resource languages, where shared representations could offer the greatest benefit. We introduce parallel tokenizers, a new framework that first trains tokenizers monolingually and then aligns their vocabularies exhaustively using bilingual dictionaries or word-to-word translation. This alignment enforces a shared semantic space across languages while naturally improving fertility balance. To assess their effectiveness, we pretrain a transformer encoder from scratch on thirteen low-resource languages and evaluate it on sentiment analysis, hate speech detection, emotion classification, and sentence embedding similarity. Across all tasks, models trained with parallel tokenizers outperform conventional multilingual baselines, confirming that rethinking tokenization is essential for advancing multilingual representation learning{--}especially in low-resource settings."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="al-kautsar-koto-2026-parallel">
<titleInfo>
<title>Parallel Tokenizers: Rethinking Vocabulary Design in Cross-Lingual Transfer of Low-Resource Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Muhammad</namePart>
<namePart type="given">Dehan</namePart>
<namePart type="family">Al Kautsar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fajri</namePart>
<namePart type="family">Koto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Tokenization forms the basis of multilingual language models, yet existing methods often limit cross-lingual transfer by mapping semantically equivalent words to different embeddings. For example, ‘I eat rice’ in English and ‘Ina cin shinkafa’ in Hausa are typically mapped to different vocabulary indices, preventing shared representations and limiting cross-lingual generalization. This problem is even more pronounced in low-resource languages, where shared representations could offer the greatest benefit. We introduce parallel tokenizers, a new framework that first trains tokenizers monolingually and then aligns their vocabularies exhaustively using bilingual dictionaries or word-to-word translation. This alignment enforces a shared semantic space across languages while naturally improving fertility balance. To assess their effectiveness, we pretrain a transformer encoder from scratch on thirteen low-resource languages and evaluate it on sentiment analysis, hate speech detection, emotion classification, and sentence embedding similarity. Across all tasks, models trained with parallel tokenizers outperform conventional multilingual baselines, confirming that rethinking tokenization is essential for advancing multilingual representation learning–especially in low-resource settings.</abstract>
<identifier type="citekey">al-kautsar-koto-2026-parallel</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.117/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>2718</start>
<end>2734</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Parallel Tokenizers: Rethinking Vocabulary Design in Cross-Lingual Transfer of Low-Resource Languages
%A Al Kautsar, Muhammad Dehan
%A Koto, Fajri
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F al-kautsar-koto-2026-parallel
%X Tokenization forms the basis of multilingual language models, yet existing methods often limit cross-lingual transfer by mapping semantically equivalent words to different embeddings. For example, ‘I eat rice’ in English and ‘Ina cin shinkafa’ in Hausa are typically mapped to different vocabulary indices, preventing shared representations and limiting cross-lingual generalization. This problem is even more pronounced in low-resource languages, where shared representations could offer the greatest benefit. We introduce parallel tokenizers, a new framework that first trains tokenizers monolingually and then aligns their vocabularies exhaustively using bilingual dictionaries or word-to-word translation. This alignment enforces a shared semantic space across languages while naturally improving fertility balance. To assess their effectiveness, we pretrain a transformer encoder from scratch on thirteen low-resource languages and evaluate it on sentiment analysis, hate speech detection, emotion classification, and sentence embedding similarity. Across all tasks, models trained with parallel tokenizers outperform conventional multilingual baselines, confirming that rethinking tokenization is essential for advancing multilingual representation learning–especially in low-resource settings.
%U https://aclanthology.org/2026.emnlp-main.117/
%P 2718-2734
Markdown (Informal)
[Parallel Tokenizers: Rethinking Vocabulary Design in Cross-Lingual Transfer of Low-Resource Languages](https://aclanthology.org/2026.emnlp-main.117/) (Al Kautsar & Koto, EMNLP 2026)
ACL