Multilingual Language Models Encode Script Over Linguistic Structure

Aastha A K Verma, Anwoy Chatterjee, Mehak Gupta, Tanmoy Chakraborty


Abstract
Multilingual language models (LMs) organize representations for typologically and orthographically diverse languages into a shared parameter space, yet the nature of this internal organization remains elusive. In this work, we investigate which linguistic properties — abstract language identity or surface-form cues — shape multilingual representations. To do so, we analyze language-associated units across different model families and scales using the Language Activation Probability Entropy (LAPE) metric, and further decompose activations with Sparse Autoencoders. We find that these units are strongly conditioned on orthography: romanization induces near-disjoint representations that align with neither native-script inputs nor English, while word-order shuffling has limited effect on unit identity. Probing shows that typological structure becomes increasingly accessible in deeper layers, while causal interventions indicate that generation is most sensitive to units that are invariant to surface-form perturbations rather than to units identified by typological alignment alone. Overall, our results suggest that multilingual LMs organize representations around surface form, with linguistic abstraction emerging gradually without collapsing into a unified interlingua.
Anthology ID:
2026.acl-long.1554
Volume:
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
July
Year:
2026
Address:
San Diego, California, United States
Editors:
Maria Liakata, Viviane P. Moreira, Jiajun Zhang, David Jurgens
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
33685–33719
Language:
URL:
https://preview.aclanthology.org/ingest-acl/2026.acl-long.1554/
DOI:
Bibkey:
Cite (ACL):
Aastha A K Verma, Anwoy Chatterjee, Mehak Gupta, and Tanmoy Chakraborty. 2026. Multilingual Language Models Encode Script Over Linguistic Structure. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 33685–33719, San Diego, California, United States. Association for Computational Linguistics.
Cite (Informal):
Multilingual Language Models Encode Script Over Linguistic Structure (Verma et al., ACL 2026)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingest-acl/2026.acl-long.1554.pdf
Checklist:
 2026.acl-long.1554.checklist.pdf