@inproceedings{cruz-etal-2026-multilinguality,
title = "Multilinguality as Sense Adaptation",
author = "Cruz, Jan Christian Blaise and
Adelani, David Ifeoluwa and
Aji, Alham Fikri",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.95/",
pages = "1951--1967",
ISBN = "979-8-89176-462-0",
abstract = "We approach multilinguality as sense adaptation: aligning latent meaning representations across languages rather than relying solely on shared parameters and scale. In this paper, we introduce SENse-based Symmetric Interlingual Alignment (SENSIA), which adapts a Backpack language model from one language to another by explicitly aligning sense-level mixtures and contextual representations on parallel data, while jointly training a target-language language modeling loss to preserve fluency. Across benchmarks on four typologically diverse languages, SENSIA generally outperforms comparable multilingual alignment methods and achieves competitive accuracy against monolingual from-scratch baselines while using 2{--}4 times less target-language data. Analyses of learned sense geometry indicate that local sense topology and global structure relative to English are largely preserved, and ablations show that the method is robust in terms of design and scale. Our results show that sense adaptation provides a simple, data-efficient, and linguistics-motivated path to multilingual fluency that is useful to low-resource languages and motivates further linguistics-motivated research."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cruz-etal-2026-multilinguality">
<titleInfo>
<title>Multilinguality as Sense Adaptation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jan</namePart>
<namePart type="given">Christian</namePart>
<namePart type="given">Blaise</namePart>
<namePart type="family">Cruz</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>
<name type="personal">
<namePart type="given">Alham</namePart>
<namePart type="given">Fikri</namePart>
<namePart type="family">Aji</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>We approach multilinguality as sense adaptation: aligning latent meaning representations across languages rather than relying solely on shared parameters and scale. In this paper, we introduce SENse-based Symmetric Interlingual Alignment (SENSIA), which adapts a Backpack language model from one language to another by explicitly aligning sense-level mixtures and contextual representations on parallel data, while jointly training a target-language language modeling loss to preserve fluency. Across benchmarks on four typologically diverse languages, SENSIA generally outperforms comparable multilingual alignment methods and achieves competitive accuracy against monolingual from-scratch baselines while using 2–4 times less target-language data. Analyses of learned sense geometry indicate that local sense topology and global structure relative to English are largely preserved, and ablations show that the method is robust in terms of design and scale. Our results show that sense adaptation provides a simple, data-efficient, and linguistics-motivated path to multilingual fluency that is useful to low-resource languages and motivates further linguistics-motivated research.</abstract>
<identifier type="citekey">cruz-etal-2026-multilinguality</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.95/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1951</start>
<end>1967</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multilinguality as Sense Adaptation
%A Cruz, Jan Christian Blaise
%A Adelani, David Ifeoluwa
%A Aji, Alham Fikri
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F cruz-etal-2026-multilinguality
%X We approach multilinguality as sense adaptation: aligning latent meaning representations across languages rather than relying solely on shared parameters and scale. In this paper, we introduce SENse-based Symmetric Interlingual Alignment (SENSIA), which adapts a Backpack language model from one language to another by explicitly aligning sense-level mixtures and contextual representations on parallel data, while jointly training a target-language language modeling loss to preserve fluency. Across benchmarks on four typologically diverse languages, SENSIA generally outperforms comparable multilingual alignment methods and achieves competitive accuracy against monolingual from-scratch baselines while using 2–4 times less target-language data. Analyses of learned sense geometry indicate that local sense topology and global structure relative to English are largely preserved, and ablations show that the method is robust in terms of design and scale. Our results show that sense adaptation provides a simple, data-efficient, and linguistics-motivated path to multilingual fluency that is useful to low-resource languages and motivates further linguistics-motivated research.
%U https://aclanthology.org/2026.findings-emnlp.95/
%P 1951-1967
Markdown (Informal)
[Multilinguality as Sense Adaptation](https://aclanthology.org/2026.findings-emnlp.95/) (Cruz et al., Findings 2026)
ACL
- Jan Christian Blaise Cruz, David Ifeoluwa Adelani, and Alham Fikri Aji. 2026. Multilinguality as Sense Adaptation. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 1951–1967, Budapest, Hungary. Association for Computational Linguistics.