How Transferable are Attribute Controllers on Pretrained Multilingual Translation Models?

Danni Liu, Jan Niehues


Abstract
Customizing machine translation models to comply with desired attributes (e.g., formality or grammatical gender) is a well-studied topic. However, most current approaches rely on (semi-)supervised data with attribute annotations. This data scarcity bottlenecks democratizing such customization possibilities to a wider range of languages, particularly lower-resource ones. This gap is out of sync with recent progress in pretrained massively multilingual translation models. In response, we transfer the attribute controlling capabilities to languages without attribute-annotated data with an NLLB-200 model as a foundation. Inspired by techniques from controllable generation, we employ a gradient-based inference-time controller to steer the pretrained model. The controller transfers well to zero-shot conditions, as it is operates on pretrained multilingual representations and is attribute- rather than language-specific. With a comprehensive comparison to finetuning-based control, we demonstrate that, despite finetuning’s clear dominance in supervised settings, the gap to inference-time control closes when moving to zero-shot conditions, especially with new and distant target languages. The latter also shows stronger domain robustness. We further show that our inference-time control complements finetuning. Moreover, a human evaluation on a real low-resource language, Bengali, confirms our findings. Our code is in the supplementary material.
Anthology ID:
2024.eacl-long.20
Volume:
Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
March
Year:
2024
Address:
St. Julian’s, Malta
Editors:
Yvette Graham, Matthew Purver
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
334–348
Language:
URL:
https://aclanthology.org/2024.eacl-long.20
DOI:
Bibkey:
Cite (ACL):
Danni Liu and Jan Niehues. 2024. How Transferable are Attribute Controllers on Pretrained Multilingual Translation Models?. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 334–348, St. Julian’s, Malta. Association for Computational Linguistics.
Cite (Informal):
How Transferable are Attribute Controllers on Pretrained Multilingual Translation Models? (Liu & Niehues, EACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.eacl-long.20.pdf
Software:
 2024.eacl-long.20.software.zip
Video:
 https://aclanthology.org/2024.eacl-long.20.mp4