mlsanityData to 14 Oct 2026

Biomedical Machine Translation for Low-Resource Arabic-Script Languages via Cross-Lingual Transfer and LoRA Adapter Merging

arXiv 2607.22300 ↗First posted 24 Jul 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

We present a systematic study of healthcare-domain cross-lingual transfer to address the scarcity of biomedical NMT resources for Arabic-script languages. We use Arabic and Persian as higher-resource pivots to improve translation for four severely low-resource targets: Dari (Afghan Persian, a standardised variety of Persian), Pashto, Sorani Kurdish (Central Kurdish, a major standardized variety of Kurdish), and Urdu (closely related to Hindi). Using LoRA fine-tuning on small decoder-only LLMs, we train textitdomain-specific pivot adapters and evaluate three transfer strategies: few-shot in-context learning, minimal supervised adaptation, and, to the best of our knowledge, for the first time in this setting, zero-data LoRA adapter merging. Supervised adaptation with just 500 sentences achieves near pivot-language quality for Dari (CHrF++ 41.01) and meaningful gains for Urdu (28.88), while adapter merging reaches within 3.5 CHrF++ of supervised adaptation for Dari at zero additional cost. Pashto and Sorani Kurdish remain insufficient for high-stakes clinical deployment exposing the limits of cross-lingual transfer when structural distance from the pivots is too great. LoRA adapter merging works surprisingly well for closely related languages, even without target-language biomedical data.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 51 references that are in the map. First posted on arXiv 24 Jul 2026.

Builds on

23 of 51 references.

References · 23

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  3. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,236
  4. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,935
  5. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,418
  6. 2014Sequence to Sequence Learning with Neural Networkscited by 2,986
  7. 2022Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference timecited by 1,396
  8. 2020COMET: A Neural Framework for MT Evaluationcited by 1,033
  9. 2023Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunchcited by 694
  10. 2016Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translationcited by 469
  11. 2019Massively Multilingual Neural Machine Translationcited by 246
  12. 2023The unreasonable effectiveness of few-shot learning for machine translationcited by 91
11 more
  1. 2020TICO-19: the Translation Initiative for Covid-19cited by 68
  2. 2024AdaMergeX: Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Mergingcited by 34
  3. 2018A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translationcited by 23
  4. 2023Physician Detection of Clinical Harm in Machine Translation: Quality Estimation Aids in Reliance and Backtranslation Identifies Critical Errorscited by 22
  5. 2022Language-Family Adapters for Low-Resource Multilingual Neural Machine Translationcited by 14
  6. 2025The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMscited by 9
  7. 2022Graphemic Normalization of the Perso-Arabic Scriptcited by 6
  8. 2025Advancing Dialectal Arabic to Modern Standard Arabic Machine Translationcited by 6
  9. 2021The interplay between language similarity and script on a novel multi-layer Algerian dialect corpuscited by 5
  10. 2025Pivot Language for Low-Resource Machine Translationcited by 3
  11. 2026Cross-Lingual Empirical Evaluation of Large Language Models for Arabic Medical Taskscited by 2

Built on by

Not cited yet.