Elamur Research⌕
☰
home / papers / 2026 / Can Linguistically Related Languages Guide LLM Translation in Low-Resource Settings?

Can Linguistically Related Languages Guide LLM Translation in Low-Resource Settings?

Summary

LLMs translate well between major languages but struggle with languages that have little digital presence. The usual remedies, large parallel corpora or fine-tuning, are out of reach for most such languages. The authors test whether a frozen LLM can be steered at inference time by pairing few-shot examples with translations into a closely related, better-resourced pivot language. From the training split they build a datastore of English–pivot–target triplets. The English side is embedded with all-MiniLM-L12-v2, and the top-k most similar triplets are retrieved by cosine similarity to serve as in-context demonstrations. Experiments cover English→Konkani (gom) and English→Tunisian Arabic (aeb) with two open ~7–8B decoder-only models, Hermes-2-Pro and TowerInstruct. They are compared against NLLB-200 and scored with BLEU and chrF++. Retrieved demonstrations improve sharply on zero-shot prompting. Adding the pivot helps only in some configurations: modestly for Konkani, and barely or not at all for Tunisian Arabic. The study is a realistic, low-compute look at when pivot-based prompting is worth using instead of fine-tuning.

Key points

  • Konkani (Eng→Gom): Hermes-2-Pro rose from 1.30 chrF++ at zero-shot to 29.62 with direct few-shot prompting and 30.34 with the pivot. This slightly exceeds NLLB-200 (26.82 chrF++, 7.51 BLEU). For TowerInstruct, the pivot raised BLEU (3.67→5.68) but lowered chrF++ (21.25→17.66).
  • Tunisian Arabic (Eng→Aeb): Hermes-2-Pro gave essentially the same chrF++ in all settings (~24.3). Its best BLEU, 6.27, came without the pivot. Both LLMs scored well above NLLB-200 on chrF++ (10.42).
  • Effect of k: Most of the benefit comes from moving from zero to a few examples. Larger k brings no further gains and often hurts, and BLEU and chrF++ tend to peak at different k. TowerInstruct degrades as prompts approach its 4K-token context window.
  • Evaluation: Scores were computed with SacreBLEU, and there was no human evaluation. The authors show Konkani outputs that are plausible in meaning but receive very low n-gram scores.
  • Limitations: The method requires a closely related high-resource pivot and good parallel data, and it is sensitive to domain mismatch in the retrieved examples.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages · Machine translation

Topics: Machine translation

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.loresmt-1.14

@inproceedings{ramasethu2026can,
  title = {{Can Linguistically Related Languages Guide LLM Translation in Low-Resource Settings?}},
  author = {Ramasethu, Aishwarya and Garg, Rohin and Allu, Niyathi and Fartale, Harshwardhan and Chan, Dun Li},
  booktitle = {Proceedings for the Ninth Workshop on Technologies for Machine Translation of Low Resource Languages (LoResMT 2026)},
  year = {2026},
  doi = {10.18653/v1/2026.loresmt-1.14},
  url = {https://doi.org/10.18653/v1/2026.loresmt-1.14}
}

References

37 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sweta Agrawal, Chunting Zhou, Mike Lewis, Luke Zettlemoyer, and Marjan Ghazvininejad. 2022. In-context examples selection for machine translation. Preprint, arXiv:2212.02437. find our card
  2. AI@Meta. 2024. Llama 3 model card. find
  3. Duarte M Alves, José Pombal, Nuno M Guerreiro, Pedro H Martins, João Alves, Amin Farajian, Ben Peters, Ricardo Rei, Patrick Fernandes, Sweta Agrawal, Pierre Colombo, José G. C. de Souza, and André F. T. Martins. 2024. Tower: An open multilingual large language model for translation-related tasks. In… find our card
  4. Naveen Arivazhagan, Ankur Bapna, Orhan Firat, Dmitry Lepikhin, Melvin Johnson, Maxim Krikun, Mia Xu Chen, Yuan Cao, George Foster, Colin Cherry, Wolfgang Macherey, Zhifeng Chen, and Yonghui Wu. 2019. Massively multilingual neural machine translation in the wild: Findings and challenges. Preprint, ar… find our card
  5. Duygu Ataman, Alexandra Birch, Nizar Habash, Marcello Federico, Philipp Koehn, and Kyunghyun Cho. 2025. Machine translation in the era of large language models:a survey of historical and emerging problems. Information, 16(9). find
  6. Houda Bouamor, Nizar Habash, and Kemal Oflazer. 2014. A multidialectal parallel corpus of arabic. In LREC, pages 1240–1245. find our card
  7. Pranjal Chitale, Jay Gala, and Raj Dabre. 2024. An empirical study of in-context learning in LLMs for machine translation. In Findings of the Association for Computational Linguistics: ACL 2024, pages 7384–7406, Bangkok, Thailand. Association for Computational Linguistics. find
  8. Monojit Choudhury. 2023. Generative AI has a language problem. Nature Human Behaviour, 7(11):1802–1803. find
  9. Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Sebastian Gehrmann, Parker Schuh, Kensen Shi, Sasha Tsvyashchenko, Joshua Maynez, Abhishek Rao, Parker Barnes, Yi Tay, Noam Shazeer, Vinodkumar Prabhakaran, and… find
  10. David M. Eberhard, Gary F. Simons, and Charles D. Fennig, editors. 2024. Ethnologue: Languages of the World, twenty-seventh edition. SIL International, Dallas, Texas. find
  11. Khalid N. Elmadani and Jan Buys. 2024. Neural machine translation between low-resource languages with synthetic pivoting. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 12144–12158, Torino, Italia. E… find
  12. Jay Gala, Pranjal A. Chitale, Raghavan AK, Varun Gumma, Sumanth Doddapaneni, Aswanth Kumar, Janki Nawale, Anupama Sujatha, Ratish Puduppully, Vivek Raghavan, Pratyush Kumar, Mitesh M. Khapra, Raj Dabre, and Anoop Kunchukuttan. 2023. Indictrans2: Towards high-quality and accessible machine translatio… find our card
  13. Xavier Garcia, Yamini Bansal, Colin Cherry, George Foster, Maxim Krikun, Melvin Johnson, and Orhan Firat. 2023. The unreasonable effectiveness of few-shot learning for machine translation. In Proceedings of the 40th International Conference on Machine Learning, volume 202 of Proceedings of Machine L… find
  14. Naman Goyal, Cynthia Gao, Vishrav Chaudhary, Peng-Jen Chen, Guillaume Wenzek, Da Ju, Sanjana Krishnan, Marc’Aurelio Ranzato, Francisco Guzmán, and Angela Fan. 2022. The flores-101 evaluation benchmark for low-resource and multilingual machine translation. Transactions of the Association for Computat… find
  15. Kenji Imamura, Masao Utiyama, and Eiichiro Sumita. 2023. Pivot translation for zero-resource language pairs based on a multilingual pretrained model. In Proceedings of Machine Translation Summit XIX, Vol. 1: Research Track, pages 348–359, Macau SAR, China. Asia-Pacific Association for Machine Transl… find
  16. Salomey Osei, Freshia Sackey, Rubungo Andre Niyongabo, Ricky Macharm, Perez Ogayo, Orevaoghene Ahia, Musie Meressa Berhe, Mofetoluwa Adeyemi, Masabata Mokgesi-Selinga, Lawrence Okegbemi, Laura Martinus, and 28 others. 2020. Participatory research for low-resourced machine translation: A case study i… find our card
  17. Albert Q Jiang, Alexandre Sablayrolles, and 1 others. 2023. Mistral 7b. arXiv preprint arXiv:2310.06825. find
  18. Wenxiang Jiao, Wenxuan Wang, Jen tse Huang, Xing Wang, Shuming Shi, and Zhaopeng Tu. 2023. Is chatgpt a good translator? yes with gpt-4 as the engine. Preprint, arXiv:2301.08745. find our card
  19. Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. 2021. The state and fate of linguistic diversity and inclusion in the nlp world. Preprint, arXiv:2004.09095. find our card
  20. OpenAI. 2024. Gpt-4.1 technical report. find
  21. Partha Pakray, Alexander Gelbukh, and Sivaji Bandyopadhyay. 2025. Natural language processing applications for low-resource languages. Natural Language Processing, 31(2):183–197. find
  22. Philipp Koehn. 2004. Statistical significance tests for machine translation evaluation. In Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing, pages 388–395. Association for Computational Linguistics. find
  23. Colin Leong, Joshua Nemecek, Jacob Mansdorfer, Anna Filighera, Abraham Owodunni, and Daniel Whitenack. 2022. Bloom library: Multimodal datasets in 300+ languages for a variety of downstream tasks. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 8608–8… find
  24. Zheng Wei Lim, Nitish Gupta, Honglin Yu, and Trevor Cohn. 2025. Multilingual fused learning for low-resource translation with llms. In International Conference on Learning Representations. find
  25. Shayne Longpre, Sneha Kudugunta, Niklas Muennighoff, I-Hung Hsu, Isaac Caswell, Alex Pentland, Sercan Ö. Arik, Chen-Yu Lee, and Sayna Ebrahimi. 2025. Atlas: Adaptive transfer scaling laws for multilingual pretraining and finetuning. Preprint, arXiv:2510.22037. — [arXiv:2510.22037](https://arxiv.org/abs/2510.22037) find
  26. Hongyuan Lu, Haoran Yang, Haoyang Huang, Dongdong Zhang, Wai Lam, and Furu Wei. 2024. Chain-of-dictionary prompting elicits translation in large language models. Preprint, arXiv:2305.06575. find
  27. Mohamed Mahdi. 2025. How well do llms understand tunisian arabic? Preprint, arXiv:2511.16683. find
  28. Niklas Muennighoff, Thomas Wang, Lintang Sutawika, Adam Roberts, Stella Biderman, Teven Le Scao, M Saiful Bari, Sheng Shen, Zheng-Xin Yong, Hailey Schoelkopf, Xiangru Tang, Dragomir Radev, Alham Fikri Aji, Khalid Almubarak, Samuel Albanie, Zaid Alyafeai, Albert Webson, Edward Raff, and Colin Raffel.… find our card
  29. Wilhelmina Nekoto, Vukosi Marivate, Tshinondiwa Matsila, Timi Fasubaa, Taiwo Fagbohungbe, Solomon Oluwole Akinola, Shamsuddeen Hassan Muhammad, Salomon Kabongo Kabenamualu, Matt Post. 2018. A call for clarity in reporting BLEU scores. In Proceedings of the Third Conference on Machine Translation: Re… find our card
  30. Ratish Puduppully, Anoop Kunchukuttan, Raj Dabre, Ai Ti Aw, and Nancy F. Chen. 2023. Decomposed prompting for machine translation between related languages using large language models. Preprint, arXiv:2305.13085. find
  31. Annie Rajan, Ambuja Salgaonkar, and Ramprasad Joshi. 2020. A survey of konkani nlp resources. Computer Science Review, 38:100299. find
  32. Abhimanyu Talwar and Julien Laasri. 2025. Pivot language for low-resource machine translation. Preprint, arXiv:2505.14553. — [arXiv:2505.14553](https://arxiv.org/abs/2505.14553) find
  33. NLLB Team. 2024. Scaling neural machine translation to 200 languages. Nature, 630:841–846. find
  34. Ryan Teknium, Jeffrey Quesnelle, and Chen Guang. 2024. Hermes 3 technical report. arXiv preprint arXiv:2408.11857. — [arXiv:2408.11857](https://arxiv.org/abs/2408.11857) find
  35. Ahmet Üstün, Viraat Aryabumi, Zheng Yong, Wei-Yin Ko, Daniel D’souza, Gbemileke Onilude, Neel Bhandari, Shivalika Singh, Hui-Lee Ooi, Amr Kayid, and 1 others. 2024. Aya model: An instruction finetuned open-access multilingual language model. In Proceedings of the 62nd Annual Meeting of the Associati… find our card
  36. David Vilar, Markus Freitag, Colin Cherry, Jiaming Luo, Viresh Ratnakar, and George Foster. 2023. Prompting PaLM for translation: Assessing strategies and performance. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15406–1542… find
  37. Zheng Xin Yong, Hailey Schoelkopf, Niklas Muennighoff, Alham Fikri Aji, David Ifeoluwa Adelani, Khalid Almubarak, M Saiful Bari, Lintang Sutawika, Jungo Kasai, Ahmed Baruwa, Genta Winata, Stella find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/ramasethu-2026-can-linguistically-related-languages-guide/