cs.CLNov 4, 2026

Familiar letters, not related languages, let unseen languages into a shared space of meaning

Authors: Yuliang Liu, Xinyu Wang, Xiang Bai

Organizations: Huazhong University of Science and Technology, Wuhan, China

Abstract

Hundreds of languages have written text but almost no translations, and whether a model learning one from untranslated text builds on a related language or familiar letters is unclear: they usually coincide. Here we show that familiar letters, not a related language, let an unseen language into a shared space of meaning. GlyphMind, a tokenizer-free vision transformer that reads text as pixels and lets any sentence be shown in any script, was pretrained to align translations across 27 writing systems. Adapted on untranslated Lao, a language and script it had never seen, it retrieved English translations seven times as often as the strongest simple baseline. Rewritten in unfamiliar letters, Lao did not enter although its close relative Thai was in pretraining; Santali, which had none, entered in Devanagari, familiar from Hindi, but not in its own script. Transliteration into familiar scripts may thus open multilingual models to languages without translations.

Figures & tables

Explore similar work

CardsList
  1. One Form to Transfer Them All: Pretraining Multilingual Language Models Beyond Native Orthography

    Aug 26, 2026Muge Zhang, Aaron Jencks, Krishna Badikela +2Language Model PretrainingMultilingual Language Models

  2. Seeing the Unseen: Visual Similarity for Pixel Language Model Adaptation

    Aug 31, 2026Ran Zhang, Miryam de Lhoneux, Wessel PoelmanContinued PretrainingLow-Resource Language Processing