Harnessing Linguistic Dissimilarity for Language Generalization on Unseen Low-Resource Varieties

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Jinju, Jung, Haeji, Roh, Youjeong, Ko, Jong Hwan, Mortensen, David R.
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914533737496576
author Kim, Jinju
Jung, Haeji
Roh, Youjeong
Ko, Jong Hwan
Mortensen, David R.
author_facet Kim, Jinju
Jung, Haeji
Roh, Youjeong
Ko, Jong Hwan
Mortensen, David R.
contents Low-resource language varieties used by specific groups remain neglected in the development of Multilingual Language Models. A great deal of cross-lingual research focuses on inter-lingual language transfer which strives to align allied varieties and minimize differences between them. However, for low-resource varieties, linguistic dissimilarity is also an important cue allowing generalization to unseen varieties. Unlike prior approaches, we propose a two-stage Language Generalization framework that focuses on capturing variety-specific cues while also exploiting rich overlap offered by high-resource source variety. First, we propose TOPPing, a source-selection method specifically designed for low-resource varieties. Second, we suggest a lightweight VACAI-Bowl architecture that learns variety-specific attributes with one branch while a parallel branch captures variety-invariant attributes using adversarial training. We evaluate our framework on structural prediction tasks, which are among the few tasks available, as proxy for performance on other downstream tasks. Using VACAI-Bowl with TOPPing yields an average 54.62% improvement in the dependency parsing task, which serves as a proxy for performance on other downstream tasks across 10 low-resource varieties.
format Preprint
id arxiv_https___arxiv_org_abs_2605_04500
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Harnessing Linguistic Dissimilarity for Language Generalization on Unseen Low-Resource Varieties
Kim, Jinju
Jung, Haeji
Roh, Youjeong
Ko, Jong Hwan
Mortensen, David R.
Computation and Language
Artificial Intelligence
Low-resource language varieties used by specific groups remain neglected in the development of Multilingual Language Models. A great deal of cross-lingual research focuses on inter-lingual language transfer which strives to align allied varieties and minimize differences between them. However, for low-resource varieties, linguistic dissimilarity is also an important cue allowing generalization to unseen varieties. Unlike prior approaches, we propose a two-stage Language Generalization framework that focuses on capturing variety-specific cues while also exploiting rich overlap offered by high-resource source variety. First, we propose TOPPing, a source-selection method specifically designed for low-resource varieties. Second, we suggest a lightweight VACAI-Bowl architecture that learns variety-specific attributes with one branch while a parallel branch captures variety-invariant attributes using adversarial training. We evaluate our framework on structural prediction tasks, which are among the few tasks available, as proxy for performance on other downstream tasks. Using VACAI-Bowl with TOPPing yields an average 54.62% improvement in the dependency parsing task, which serves as a proxy for performance on other downstream tasks across 10 low-resource varieties.
title Harnessing Linguistic Dissimilarity for Language Generalization on Unseen Low-Resource Varieties
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2605.04500