Multilingual Pretraining for Pixel Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kesen, Ilker, Lotz, Jonas F., Ziegler, Ingo, Rust, Phillip, Elliott, Desmond
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912741970673664
author Kesen, Ilker
Lotz, Jonas F.
Ziegler, Ingo
Rust, Phillip
Elliott, Desmond
author_facet Kesen, Ilker
Lotz, Jonas F.
Ziegler, Ingo
Rust, Phillip
Elliott, Desmond
contents Pixel language models operate directly on images of rendered text, eliminating the need for a fixed vocabulary. While these models have demonstrated strong capabilities for downstream cross-lingual transfer, multilingual pretraining remains underexplored. We introduce PIXEL-M4, a model pretrained on four visually and linguistically diverse languages: English, Hindi, Ukrainian, and Simplified Chinese. Multilingual evaluations on semantic and syntactic tasks show that PIXEL-M4 outperforms an English-only counterpart on non-Latin scripts. Word-level probing analyses confirm that PIXEL-M4 captures rich linguistic features, even in languages not seen during pretraining. Furthermore, an analysis of its hidden representations shows that multilingual pretraining yields a semantic embedding space closely aligned across the languages used for pretraining. This work demonstrates that multilingual pretraining substantially enhances the capability of pixel language models to effectively support a diverse set of languages.
format Preprint
id arxiv_https___arxiv_org_abs_2505_21265
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multilingual Pretraining for Pixel Language Models
Kesen, Ilker
Lotz, Jonas F.
Ziegler, Ingo
Rust, Phillip
Elliott, Desmond
Computation and Language
Artificial Intelligence
Pixel language models operate directly on images of rendered text, eliminating the need for a fixed vocabulary. While these models have demonstrated strong capabilities for downstream cross-lingual transfer, multilingual pretraining remains underexplored. We introduce PIXEL-M4, a model pretrained on four visually and linguistically diverse languages: English, Hindi, Ukrainian, and Simplified Chinese. Multilingual evaluations on semantic and syntactic tasks show that PIXEL-M4 outperforms an English-only counterpart on non-Latin scripts. Word-level probing analyses confirm that PIXEL-M4 captures rich linguistic features, even in languages not seen during pretraining. Furthermore, an analysis of its hidden representations shows that multilingual pretraining yields a semantic embedding space closely aligned across the languages used for pretraining. This work demonstrates that multilingual pretraining substantially enhances the capability of pixel language models to effectively support a diverse set of languages.
title Multilingual Pretraining for Pixel Language Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2505.21265