Pith. sign in

REVIEW 2 major objections 5 minor 104 references

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

T0 review · 2 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Zero-data LoRA adapter merging of Arabic and Persian biomedical adapters reaches within 3.5 CHrF++ points of supervised adaptation for Dari on the dev set and within 1.36 points on the test set, while supervised adaptation with just 500 sen

desk verdict A systematic, honestly-written empirical study of adapter merging for biomedical MT into four Arabic-script low-resource languages; the headline gaps are plausible but rest on single-seed runs, so the quantitative claims need variance reporting before they hold weight. read the letter →

arxiv 2607.22300 v1 pith:C3QPMYQI submitted 2026-07-24 cs.CL

Abdullah Alabdullah , Arash Eslamighayour , Sarp Harbalioglu , Lifeng Han This is my paper
classification cs.CL
keywords biomedicalmachinetranslationcross-lingualtransferLoRAadaptermerginglow-resourcelanguagesArabicscriptDarizero-dataadaptationCHrF++evaluation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Biomedical translation is a lives-at-stake task, yet four Arabic-script languages—Dari, Pashto, Sorani Kurdish, and Urdu—have almost no healthcare-domain parallel data. The paper tries to establish that knowledge encoded in Arabic and Persian biomedical LoRA adapters can be transferred to these targets without any target-language biomedical data. Its central finding is that zero-data adapter merging comes within 3.5 CHrF++ points of supervised adaptation for Dari on the dev set and within 1.36 on the test set, while supervised adaptation with just 500 general-domain sentences matches pivot-level quality for Dari and gives meaningful Urdu gains. Sympathetic readers should care because, if true, this offers a nearly free path to initial biomedical MT for closely related low-resource languages, and it sharply exposes where transfer stops working (Pashto, Sorani).

What carries the argument

LoRA adapters: rank-16 low-rank matrix updates on a frozen decoder-only LLM, so the only learned parameters per language are small adapter matrices; because both adapters share the same frozen base, their differences from the base form task vectors that live in the same parameter space. Four merging functions (simple/weighted averaging, TIES-Merging, DARE) combine the Arabic and Persian task vectors to produce a zero-data multilingual adapter; the merging function does the work of redirecting biomedical knowledge to target languages.

What would settle it

Have a qualified translator review 200 Dari sentences produced by the merged adapter and by the supervised-adaptation model. If the merged outputs carry a clinically significant error rate no better than the zero-shot baseline, despite the 1.36-point CHrF++ gap, the claim that zero-data merging approaches supervised quality for clinical use collapses.

Watch

Extended reading notes

Core claim

The paper demonstrates that a merged LoRA adapter, formed by tensor arithmetic over the Arabic and Persian task vectors from a shared frozen base model, yields Dari translations within 3.5 CHrF++ (dev) and 1.36 (test) of supervised adaptation—without any Dari training data—and that fine-tuning either pivot adapter on only 500 FLORES-200 sentences produces near-pivot quality for Dari (41.01) and meaningful quality for Urdu (28.88). It further finds a model inversion: the weaker pivot model transfers better than the stronger one, suggesting a trade-off between pivot specialisation and cross-lingual generalisation.

Load-bearing premise

The central premise is that 500 general-domain sentences can redirect biomedical knowledge already stored in the pivot adapters, and that CHrF++ gains reflect translation adequacy in a medical setting; if either fails, the headline numbers overstate deployable quality.

Editorial extensions

If this is right

  • Zero-data adapter merging is a viable starting point for closely related low-resource languages when no target-language data exists.
  • 500 sentences of general-domain data are enough to reach near-pivot quality for Dari, so minimal supervision can replace large in-domain corpora for close relatives.
  • Pivot choice becomes irrelevant once target-language supervision is available; the Arabic and Persian adapters converge within 1.0 CHrF++ after adaptation.
  • Pashto and Sorani Kurdish cannot be brought to clinically usable quality by cross-lingual transfer alone; in-domain data or a closer pivot is required.
  • The model inversion result implies that choosing the best pivot model for the pivot language is not the same as choosing the best model for transfer.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the same merging recipe could be tested on other script-sharing families (e.g., Hindi–Urdu, or Spanish–Portuguese biomedical adapters) to see whether the close-relative result generalises.
  • Editorial inference: because the adaptation data is general-domain, replacing it with even 200 in-domain biomedical sentences may push Pashto and Sorani over the usability threshold; this is a direct testable extension.
  • Editorial inference: the CHrF++ metric rewards surface similarity, so the headline 1.36-point gap may not reflect clinical adequacy; a small human review of Dari outputs is the natural next check.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper studies cross-lingual biomedical MT from English into four low-resource Arabic-script languages (Dari, Pashto, Sorani Kurdish, Urdu) using Arabic and Persian as pivots. The authors train LoRA adapters on 25k biomedical sentence pairs for en→ar and en→fa, then compare three transfer strategies: few-shot in-context learning, minimal supervised adaptation on 500 general-domain FLORES-200 sentences, and zero-data merging of the two pivot adapters via simple/weighted averaging, TIES, and DARE. Evaluation is on the held-out TICO-19 dev/test sets with CHrF++. The central claims are that supervised adaptation reaches near-pivot quality for Dari (41.01 dev, 36.95–38.52 test) and meaningful quality for Urdu (28.88), while zero-data adapter merging comes within 3.5 CHrF++ (dev) and 1.36 CHrF++ (test) of supervised adaptation for Dari, and that Pashto/Sorani remain far from usable. A secondary finding is an apparent model inversion where Llama 3.2 3B transfers better than Gemma 2 2B despite weaker pivot performance.

Significance. If the results are robust, the paper makes a useful practical contribution: it demonstrates a computationally cheap, zero-target-data route to biomedical MT for a closely related language cluster, and it does so with a clean evaluation protocol that keeps TICO-19 entirely out of training and uses an external benchmark for both model selection and final testing. The comparison of adapter merging against joint multilingual training and supervised adaptation is informative, and the candid limitations section correctly identifies most residual confounds. The main risk is that all headline differences—especially the 1.36–3.5 CHrF++ margins that define the merging claim—are based on a single training run per configuration, so the paper's quantitative contribution is not yet statistically grounded. The work is within the scope of the journal and, after the variance and reporting issues are addressed, could be an accept.

major comments (2)
  1. [§B.1, Tables 2–3, §5.5–5.6] All headline quantitative claims rest on a single run with a fixed seed (B.1: 'A fixed random seed of 3407 is used throughout'). The abstract's 'within 3.5 CHrF++' (dev) and §5.6's 'within 1.36 points' (test) are statements about the size of a difference, but no variance estimates, confidence intervals, or significance tests are provided. The instability of the comparison across splits reinforces the concern: on dev, the best Dari adaptation is AR-initialized (41.01) and on test it is FA-initialized (38.52), while AR-initialized test adaptation (36.95) falls below the simple-average merge score (37.16). This pattern is consistent with high run-to-run variability. I request multiple seeds (at least 3–5) for the headline configurations, with means and bootstrap CIs or paired tests, so the 1.36–3.5 CHrF++ margins are interpretable.
  2. [Table 3 caption, §5.6] The caption says 'Stage 4 uses Llama 3.2 3B; all other stages use Gemma 2 2B.' If 'Stage 4' refers to adaptation fine-tuning (§B.4), then the test-set adaptation scores in Table 3 (AR 36.95, FA 38.52) are Llama, while the merge rows are Gemma. The sentence in §5.6 'within 1.36 points of supervised adaptation' would then compare a Gemma merge against a Llama adaptation, which is not an apples-to-apples comparison. If all adaptation and merge rows in Table 3 are actually Gemma, the caption must be corrected and the notion of 'stage' defined. Please clarify the base model for every row/column in Tables 2–3 and ensure that any merge-vs-adaptation comparison holds the base model fixed.
minor comments (5)
  1. [Tables 2–3] Table 2 appears to report development-set scores but the caption never says 'dev'; Table 3 is the test set but only §5.6 identifies it as such. Add explicit 'development' / 'test' labels to the captions and in the running text.
  2. [§5.3] The model-inversion claim cites specific numbers (Pashto 16.70 vs. 6.28; Urdu 22.26 vs. 9.96) that do not all appear in Table 2, which shows only the Llama AR+FA few-shot condition. Report the full few-shot matrix for both base models so the claimed inversion is verifiable.
  3. [§5.5] The sentence 'Table 1 shows the full comparison' should presumably refer to Table 2; Table 1 is the linguistic-difficulty table. Please correct the cross-reference.
  4. [§4.2.2, §4.3] Text normalization is applied to Arabic, Persian, and Dari but not to Pashto, Sorani Kurdish, or Urdu, because no standardized tools exist. Table 4 draws cross-language conclusions from CHrF++ scores that are therefore not strictly comparable across languages. Please add a sensitivity analysis or an explicit caveat that CHrF++ differences across languages may partly reflect normalization asymmetry.
  5. [§1 and §B.3] Minor wording issues: the citation '(Talwar and Laasri, 2025)' is inserted mid-sentence after 'pivot languages' in an awkward place; and §B.3 has a typo, 'monolingual dapters'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: all central claims are empirical evaluations against the external TICO-19 benchmark; no fitted quantity is recycled as evidence.

full rationale

The paper's derivation chain is empirical rather than definitional. Pivot adapters are trained on external biomedical corpora (PEACH, Esposito), transfer strategies are defined by explicit tensor arithmetic (Eqs. 2–5), and both adaptation and merging results are measured against the held-out TICO-19 benchmark. The headline 'within 3.5 CHrF++' and test-set 'within 1.36 points' are observed comparisons, not consequences of how the merged adapter is constructed; simple averaging (α=0.5) is not fitted to the target language. Model and α selection on the TICO-19 dev split is standard practice and is then confirmed on the 2,100-sentence test set. The self-citations present (Alabdullah et al. 2025 for metric choice; Han et al. for domain motivation) are not load-bearing: the metric choice is independently motivated and the transfer result is not justified by these citations. The Limitations section honestly flags metric limitations and confounding factors ('automatic metrics such as BLEU and CHRF++ may not fully capture semantic adequacy'; 'some performance differences attributed to cross-lingual transfer may also partially reflect underlying dataset characteristics'); these are validity caveats, not circularity. Concerns about single-seed runs and lack of significance testing are statistical robustness issues, not evidence that any prediction reduces to its inputs. The paper is self-contained against an external benchmark, so the appropriate circularity score is 0.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claims rest mainly on domain assumptions about evaluation validity and data representativeness rather than on mathematical axioms; no new entities are introduced. The most important free parameters are the merge weights and the hand-fixed data budgets and hyperparameters.

free parameters (5)
  • Weighted-average coefficient α (merge) = 0.3 (Persian-biased), 0.7 (Arabic-biased)
    Two α values are tested and the best per target language is reported; the claim that optimal merging tracks linguistic proximity is derived from these selected values, not from a fixed rule.
  • TIES density k = 0.2
    Hand-chosen; no sensitivity analysis is provided.
  • DARE drop rate p = 0.3
    Hand-chosen; no sensitivity analysis is provided.
  • LoRA rank r and scale α = r=16, α=16 (monolingual); r=32, α=64 (joint)
    Hand-chosen standard configuration; central results could depend on adapter capacity.
  • Target-language supervision size = 500 FLORES-200 sentences per language
    Hand-chosen data budget; the claim that '500 sentences suffice' would need a scaling curve to be fully supported.
assumptions (5)
  • domain assumption TICO-19 references and CHrF++ are adequate proxies for biomedical translation adequacy.
    Used as the sole evaluation; the authors acknowledge in Limitations that 'automatic metrics such as BLEU and CHRF++ may not fully capture semantic adequacy' and that human evaluation is future work.
  • domain assumption General-domain FLORES-200 target sentences can redirect biomedical knowledge from pivot adapters.
    Adaptation fine-tuning uses 500 general-domain sentences per target; if in-domain data is required, the reported transfer results for Dari/Urdu would not reflect biomedical capability. The authors flag domain mismatch in Limitations.
  • standard math Arabic and Persian LoRA task vectors trained from the same frozen base are compatible for element-wise merging.
    This is the standard task-arithmetic assumption (Ilharco et al.; Wortsman et al.) and holds by construction for same-base same-rank adapters, but it is still an assumption about parameter-space geometry.
  • domain assumption FLORES-200 demonstration/adaptation data does not overlap with TICO-19 test data.
    The paper asserts 'TICO-19 exclusively for evaluation and testing, ensuring no data contamination' without providing a formal overlap check.
  • domain assumption PEACH and Esposito are high-quality enough to encode biomedical domain knowledge.
    PEACH is manually aligned; Esposito is auto-aligned with 82% human-evaluated good/very good alignment. The paper relies on this without independent verification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Biomedical Machine Translation for Low-Resource Arabic-Script Languages via Cross-Lingual Transfer and LoRA Adapter Merging." pith.science (2026). https://pith.science/paper/C3QPMYQI

@misc{pith2026260722300,
  author       = {Pith},
  title        = {Pith review of: Biomedical Machine Translation for Low-Resource Arabic-Script Languages via Cross-Lingual Transfer and LoRA Adapter Merging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C3QPMYQI}},
  note         = {Machine review of arXiv:2607.22300}
}
read the original abstract

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 \textbf{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 \textit{domain-specific pivot adapters} and evaluate \textbf{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.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

104 extracted references · 9 canonical work pages

  1. [1]

    arXiv preprint arXiv:2507.20301 , year=

    Advancing dialectal Arabic to modern standard Arabic machine translation , author=. arXiv preprint arXiv:2507.20301 , year=

    arXiv

  2. [2]

    2022 , eprint=

    No Language Left Behind: Scaling Human-Centered Machine Translation , author=. 2022 , eprint=

    2022

  3. [3]

    Investigating Massive Multilingual Pre-Trained Machine Translation Models for Clinical Domain via Transfer Learning

    Han, Lifeng and Erofeev, Gleb and Sorokina, Irina and Gladkoff, Serge and Nenadic, Goran. Investigating Massive Multilingual Pre-Trained Machine Translation Models for Clinical Domain via Transfer Learning. Proceedings of the 5th Clinical Natural Language Processing Workshop. 2023. doi:10.18653/v1/2023.clinicalnlp-1.5

    doi:10.18653/v1/2023.clinicalnlp-1.5 2023

  4. [4]

    BMJ Qual Saf

    Evaluation of the accuracy and safety of machine translation of patient-specific discharge instructions: a comparative analysis. BMJ Qual Saf. 2026; 35 (3): 150--8 , author=

    2026

  5. [5]

    Frontiers in Digital Health , VOLUME=

    Han, Lifeng and Gladkoff, Serge and Erofeev, Gleb and Sorokina, Irina and Galiano, Betty and Nenadic, Goran , TITLE=. Frontiers in Digital Health , VOLUME=. 2024 , URL=. doi:10.3389/fdgth.2024.1211564 , ISSN=

    arXiv 2024

  6. [6]

    IETE Journal of Research , volume =

    Shweta Chauhan and Philemon Daniel and Archita Mishra and Abhay Kumar , title =. IETE Journal of Research , volume =. 2023 , publisher =. doi:10.1080/03772063.2021.1962745 , URL =

    arXiv 2023

  7. [7]

    Findings of the A mericas NLP 2025 Shared Tasks on Machine Translation, Creation of Educational Material, and Translation Metrics for Indigenous Languages of the A mericas

    De Gibert, Ona and Pugh, Robert and Marashian, Ali and Vazquez, Raul and Ebrahimi, Abteen and Denisov, Pavel and Rice, Enora and Gow-Smith, Edward and Prieto, Juan and Robles, Melissa and Manrique, Rub. Findings of the A mericas NLP 2025 Shared Tasks on Machine Translation, Creation of Educational Material, and Translation Metrics for Indigenous Languages...

    doi:10.18653/v1/2025 2025

  8. [8]

    Transfer Learning for Low-Resource Neural Machine Translation

    Zoph, Barret and Yuret, Deniz and May, Jonathan and Knight, Kevin. Transfer Learning for Low-Resource Neural Machine Translation. Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. 2016. doi:10.18653/v1/D16-1163

    doi:10.18653/v1/d16-1163 2016

Show all 104 references
  1. [9]

    Journal of Machine Learning Research , year =

    Angela Fan and Shruti Bhosale and Holger Schwenk and Zhiyi Ma and Ahmed El-Kishky and Siddharth Goyal and Mandeep Baines and Onur Celebi and Guillaume Wenzek and Vishrav Chaudhary and Naman Goyal and Tom Birch and Vitaliy Liptchinsky and Sergey Edunov and Michael Auli and Arma...

  2. [10]

    arXiv preprint arXiv:1409.0473 , year=

    Neural machine translation by jointly learning to align and translate , author=. arXiv preprint arXiv:1409.0473 , year=

    arXiv

  3. [11]

    The Twelfth International Conference on Learning Representations , year=

    A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models , author=. The Twelfth International Conference on Learning Representations , year=

  4. [12]

    Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis

    Zhu, Wenhao and Liu, Hongyi and Dong, Qingxiu and Xu, Jingjing and Huang, Shujian and Kong, Lingpeng and Chen, Jiajun and Li, Lei. Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis. Findings of the Association for Computational Linguis...

    2024 doi

  5. [13]

    Comparing Large Language Models and Traditional Machine Translation Tools for Translating Medical Consultation Summaries: Quantitative Pilot Feasibility Study

    Li, Andy and Zhou, Wei and Hoda, Rashina and Bain, Chris and Poon, Peter. Comparing Large Language Models and Traditional Machine Translation Tools for Translating Medical Consultation Summaries: Quantitative Pilot Feasibility Study. JMIR Form Res. 2026. doi:10.2196/85169

    2026 doi

  6. [14]

    Edward J Hu and yelong shen and Phillip Wallis and Zeyuan Allen-Zhu and Yuanzhi Li and Shean Wang and Lu Wang and Weizhu Chen , booktitle=. Lo. 2022 , url=

    2022

  7. [15]

    Tuning LLM s with Contrastive Alignment Instructions for Machine Translation in Unseen, Low-resource Languages

    Mao, Zhuoyuan and Yu, Yen. Tuning LLM s with Contrastive Alignment Instructions for Machine Translation in Unseen, Low-resource Languages. Proceedings of the Seventh Workshop on Technologies for Machine Translation of Low-Resource Languages (LoResMT 2024). 2024. doi:10.18653/v...

    2024 doi

  8. [16]

    Toward Low-Resource Languages Machine Translation: A Language-Specific Fine-Tuning With LoRA for Specialized Large Language Models , year=

    Liang, Xiao and Jasmina Khaw, Yen-Min and Liew, Soung-Yue and Tan, Tien-Ping and Qin, Donghong , journal=. Toward Low-Resource Languages Machine Translation: A Language-Specific Fine-Tuning With LoRA for Specialized Large Language Models , year=

  9. [17]

    2023 , url=

    Tim Dettmers and Artidoro Pagnoni and Ari Holtzman and Luke Zettlemoyer , booktitle=. 2023 , url=

    2023

  10. [18]

    Massively Multilingual Neural Machine Translation

    Aharoni, Roee and Johnson, Melvin and Firat, Orhan. Massively Multilingual Neural Machine Translation. Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Pape...

    2019 doi

  11. [19]

    The interplay between language similarity and script on a novel multi-layer A lgerian dialect corpus

    Touileb, Samia and Barnes, Jeremy. The interplay between language similarity and script on a novel multi-layer A lgerian dialect corpus. Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. 2021. doi:10.18653/v1/2021.findings-acl.324

    2021 doi

  12. [20]

    Proceedings of the 39th International Conference on Machine Learning , pages =

    Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time , author =. Proceedings of the 39th International Conference on Machine Learning , pages =. 2022 , editor =

    2022

  13. [21]

    The Eleventh International Conference on Learning Representations , year=

    Editing models with task arithmetic , author=. The Eleventh International Conference on Learning Representations , year=

  14. [22]

    TIES-Merging: Resolving Interference When Merging Models , volume =

    Yadav, Prateek and Tam, Derek and Choshen, Leshem and Raffel, Colin and Bansal, Mohit , booktitle =. TIES-Merging: Resolving Interference When Merging Models , volume =

  15. [23]

    Forty-first International Conference on Machine Learning , year=

    Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch , author=. Forty-first International Conference on Machine Learning , year=

  16. [24]

    Neural Machine Translation on scarce-resource condition: A case-study on Persian-English , year=

    Bastan, Mohaddeseh and Khadivi, Shahram and Homayounpour, Mohammad Mehdi , booktitle=. Neural Machine Translation on scarce-resource condition: A case-study on Persian-English , year=

  17. [25]

    Iranian Journal of Applied Language Studies , volume =

    Zand Rahimi, Mina and Madayenzadeh, Moein and Alizadeh, Mahdi , title =. Iranian Journal of Applied Language Studies , volume =. 2017 , publisher =. doi:10.22111/ijals.2017.4233 , abstract =

    2017

  18. [26]

    A da M erge X : Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Merging

    Zhao, Yiran and Zhang, Wenxuan and Wang, Huiming and Kawaguchi, Kenji and Bing, Lidong. A da M erge X : Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Merging. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association ...

    2025 doi

  19. [27]

    The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLM s

    Bandarkar, Lucas and Peng, Nanyun. The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLM s. Proceedings of the 5th Workshop on Multilingual Representation Learning (MRL 2025). 2025. doi:10.18653/v1/2025.mrl-main.10

    2025 doi

  20. [28]

    Language-Family Adapters for Low-Resource Multilingual Neural Machine Translation

    Chronopoulou, Alexandra and Stojanovski, Dario and Fraser, Alexander. Language-Family Adapters for Low-Resource Multilingual Neural Machine Translation. Proceedings of the Sixth Workshop on Technologies for Machine Translation of Low-Resource Languages (LoResMT 2023). 2023. do...

    2023 doi

  21. [29]

    Unlocking the Potential of Model Merging for Low-Resource Languages

    Tao, Mingxu and Zhang, Chen and Huang, Quzhe and Ma, Tianyao and Huang, Songfang and Zhao, Dongyan and Feng, Yansong. Unlocking the Potential of Model Merging for Low-Resource Languages. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024. doi:10.18653/...

    2024 doi

  22. [30]

    arXiv preprint arXiv:2506.00250 , year=

    PersianMedQA: Evaluating large language models on a persian-english bilingual medical question answering benchmark , author=. arXiv preprint arXiv:2506.00250 , year=

    arXiv

  23. [31]

    Al-Kabi and Taghreed M

    Mohammed N. Al-Kabi and Taghreed M. Hailat and Emad M. Al-Shawakfa and Izzat M. Alsmadi , doi =. Evaluating English to Arabic Machine Translation Using BLEU , journal =. 2013 , publisher =

    2013

  24. [32]

    Computer Speech & Language , volume=

    Adopting machine translation in the healthcare sector: A methodological multi-criteria review , author=. Computer Speech & Language , volume=. 2024 , publisher=

    2024

  25. [33]

    PeerJ Computer Science , volume=

    TurkMedNLI: a Turkish medical natural language inference dataset through large language model based translation , author=. PeerJ Computer Science , volume=. 2025 , publisher=

    2025

  26. [34]

    Findings of the WMT 2022 Biomedical Translation Shared Task: Monolingual Clinical Case Reports

    Neves, Mariana and Jimeno Yepes, Antonio and Siu, Amy and Roller, Roland and Thomas, Philippe and Vicente Navarro, Maika and Yeganova, Lana and Wiemann, Dina and Di Nunzio, Giorgio Maria and Vezzani, Federica and Gerardin, Christel and Bawden, Rachel and Estrada, Darryl Johan ...

    2022

  27. [35]

    Findings of the WMT 2024 Biomedical Translation Shared Task: Test Sets on Abstract Level

    Neves, Mariana and Grozea, Cristian and Thomas, Philippe and Roller, Roland and Bawden, Rachel and N \'e v \'e ol, Aur \'e lie and Castle, Steffen and Bonato, Vanessa and Di Nunzio, Giorgio Maria and Vezzani, Federica and Vicente Navarro, Maika and Yeganova, Lana and Jimeno Ye...

    2024 doi

  28. [36]

    Physician Detection of Clinical Harm in Machine Translation: Quality Estimation Aids in Reliance and Backtranslation Identifies Critical Errors

    Mehandru, Nikita and Agrawal, Sweta and Xiao, Yimin and Gao, Ge and Khoong, Elaine and Carpuat, Marine and Salehi, Niloufar. Physician Detection of Clinical Harm in Machine Translation: Quality Estimation Aids in Reliance and Backtranslation Identifies Critical Errors. Proceed...

    2023 doi

  29. [37]

    Junczys-Dowmunt, Marcin and Grundkiewicz, Roman and Dwojak, Tomasz and Hoang, Hieu and Heafield, Kenneth and Neckermann, Tom and Seide, Frank and Germann, Ulrich and Aji, Alham Fikri and Bogoychev, Nikolay and Martins, Andr \'e F. T. and Birch, Alexandra. M arian: Fast Neural ...

    2018 doi

  30. [38]

    F acebook AI ' s WMT 21 News Translation Task Submission

    Tran, Chau and Bhosale, Shruti and Cross, James and Koehn, Philipp and Edunov, Sergey and Fan, Angela. F acebook AI ' s WMT 21 News Translation Task Submission. Proceedings of the Sixth Conference on Machine Translation. 2021

    2021

  31. [39]

    Journal of Artificial Intelligence Research , volume=

    Domain adaptation and multi-domain adaptation for neural machine translation: A survey , author=. Journal of Artificial Intelligence Research , volume=

  32. [40]

    Domain Adaptation of Neural Machine Translation by Lexicon Induction

    Hu, Junjie and Xia, Mengzhou and Neubig, Graham and Carbonell, Jaime. Domain Adaptation of Neural Machine Translation by Lexicon Induction. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. 2019. doi:10.18653/v1/P19-1286

    2019 doi

  33. [41]

    Curriculum Learning for Domain Adaptation in Neural Machine Translation

    Zhang, Xuan and Shapiro, Pamela and Kumar, Gaurav and McNamee, Paul and Carpuat, Marine and Duh, Kevin. Curriculum Learning for Domain Adaptation in Neural Machine Translation. Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computation...

    2019 doi

  34. [42]

    Rapid Adaptation of Neural Machine Translation to New Languages

    Neubig, Graham and Hu, Junjie. Rapid Adaptation of Neural Machine Translation to New Languages. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. 2018. doi:10.18653/v1/D18-1103

    2018 doi

  35. [43]

    A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation

    Lakew, Surafel Melaku and Cettolo, Mauro and Federico, Marcello. A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation. Proceedings of the 27th International Conference on Computational Linguistics. 2018

    2018

  36. [44]

    and Krikun, Maxim and Wu, Yonghui and Chen, Zhifeng and Thorat, Nikhil and Vi \'e gas, Fernanda and Wattenberg, Martin and Corrado, Greg and Hughes, Macduff and Dean, Jeffrey

    Johnson, Melvin and Schuster, Mike and Le, Quoc V. and Krikun, Maxim and Wu, Yonghui and Chen, Zhifeng and Thorat, Nikhil and Vi \'e gas, Fernanda and Wattenberg, Martin and Corrado, Greg and Hughes, Macduff and Dean, Jeffrey. G oogle ' s Multilingual Neural Machine Translatio...

    2017 doi

  37. [45]

    Transactions of the Association for Computational Linguistics , volume =

    Khashabi, Daniel and Cohan, Arman and Shakeri, Siamak and Hosseini, Pedram and Pezeshkpour, Pouya and Alikhani, Malihe and Aminnaseri, Moin and Bitaab, Marzieh and Brahman, Faeze and Ghazarian, Sarik and Gheini, Mozhdeh and Kabiri, Arman and Mahabagdi, Rabeeh Karimi and Memarr...

    2021

  38. [46]

    Esposito: An E nglish- P ersian Scientific Parallel Corpus for Machine Translation

    Esalati, Mersad and Dousti, Mohammad Javad and Faili, Heshaam. Esposito: An E nglish- P ersian Scientific Parallel Corpus for Machine Translation. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLIN...

    2024

  39. [47]

    Daniel Han, Michael Han and Unsloth team , title =

  40. [48]

    ARBERT & MARBERT : Deep Bidirectional Transformers for A rabic

    Abdul-Mageed, Muhammad and Elmadany, AbdelRahim and Nagoudi, El Moatez Billah. ARBERT & MARBERT : Deep Bidirectional Transformers for A rabic. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on...

    2021 doi

  41. [49]

    Corpora , volume =

    Al-Sabbagh, Rania , title =. Corpora , volume =. 2024 , doi =. https://doi.org/10.3366/cor.2024.0320 , abstract =

    2024

  42. [50]

    Exploring Segmentation Approaches for Neural Machine Translation of Code-Switched E gyptian A rabic- E nglish Text

    Gaser, Marwa and Mager, Manuel and Hamed, Injy and Habash, Nizar and Abdennadher, Slim and Vu, Ngoc Thang. Exploring Segmentation Approaches for Neural Machine Translation of Code-Switched E gyptian A rabic- E nglish Text. Proceedings of the 17th Conference of the European Cha...

    2023 doi

  43. [51]

    Journal of Healthcare Informatics Research , volume=

    BioBERTurk: Exploring Turkish biomedical language model development strategies in low-resource setting , author=. Journal of Healthcare Informatics Research , volume=. 2023 , publisher=

    2023

  44. [52]

    2017 , eprint=

    Linguistically Motivated Vocabulary Reduction for Neural Machine Translation from Turkish to English , author=. 2017 , eprint=

    2017

  45. [53]

    Developing Question-Answering Models in Low-Resource Languages: A Case Study on Turkish Medical Texts Using Transformer-Based Approaches , year=

    İncidelen, Mert and Aydoğan, Murat , booktitle=. Developing Question-Answering Models in Low-Resource Languages: A Case Study on Turkish Medical Texts Using Transformer-Based Approaches , year=

  46. [54]

    Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains

    Zouhar, Vil \'e m and Ding, Shuoyang and Currey, Anna and Badeka, Tatyana and Wang, Jenyuan and Thompson, Brian. Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2...

    2024 doi

  47. [55]

    Are LLM s Breaking MT Metrics? Results of the WMT 24 Metrics Shared Task

    Freitag, Markus and Mathur, Nitika and Deutsch, Daniel and Lo, Chi-Kiu and Avramidis, Eleftherios and Rei, Ricardo and Thompson, Brian and Blain, Frederic and Kocmi, Tom and Wang, Jiayi and Adelani, David Ifeoluwa and Buchicchio, Marianna and Zerva, Chrysoula and Lavie, Alon. ...

    2024 doi

  48. [56]

    Proceedings of the Seventh Conference on Machine Translation (WMT) , pages=

    Results of WMT22 metrics shared task: Stop using BLEU--neural metrics are better and more robust , author=. Proceedings of the Seventh Conference on Machine Translation (WMT) , pages=

  49. [57]

    TICO -19: the Translation Initiative for CO vid-19

    Anastasopoulos, Antonios and Cattelan, Alessandro and Dou, Zi-Yi and Federico, Marcello and Federmann, Christian and Genzel, Dmitriy and Guzm. TICO -19: the Translation Initiative for CO vid-19. Proceedings of the 1st Workshop on NLP for COVID -19 (Part 2) at EMNLP 2020. 2020....

    2020 doi

  50. [58]

    O pen WHO : A Document-Level Parallel Corpus for Health Translation in Low-Resource Languages

    Merx, Raphael and Suominen, Hanna and Cohn, Trevor and Vylomova, Ekaterina. O pen WHO : A Document-Level Parallel Corpus for Health Translation in Low-Resource Languages. Proceedings of the Tenth Conference on Machine Translation. 2025. doi:10.18653/v1/2025.wmt-1.8

    2025 doi

  51. [59]

    2022 , eprint=

    Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model , author=. 2022 , eprint=

    2022

  52. [60]

    Language Models are Few-Shot Learners , volume =

    Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie and Kaplan, Jared D and Dhariwal, Prafulla and Neelakantan, Arvind and Shyam, Pranav and Sastry, Girish and Askell, Amanda and Agarwal, Sandhini and Herbert-Voss, Ariel and Krueger, Gretchen and Henighan, Tom a...

  53. [61]

    2024 , eprint=

    Gemma 2: Improving Open Language Models at a Practical Size , author=. 2024 , eprint=

    2024

  54. [62]

    2024 , eprint=

    The Llama 3 Herd of Models , author=. 2024 , eprint=

    2024

  55. [63]

    2023 , eprint=

    LLaMA: Open and Efficient Foundation Language Models , author=. 2023 , eprint=

    2023

  56. [64]

    2023 , eprint=

    Llama 2: Open Foundation and Fine-Tuned Chat Models , author=. 2023 , eprint=

    2023

  57. [65]

    RoFormer: Enhanced transformer with Rotary Position Embedding , journal =

    Jianlin Su and Murtadha Ahmed and Yu Lu and Shengfeng Pan and Wen Bo and Yunfeng Liu , keywords =. RoFormer: Enhanced transformer with Rotary Position Embedding , journal =. 2024 , issn =. doi:https://doi.org/10.1016/j.neucom.2023.127063 , url =

    2024

  58. [66]

    2025 , eprint=

    Qwen2.5 Technical Report , author=. 2025 , eprint=

    2025

  59. [67]

    COMET : A Neural Framework for MT Evaluation

    Rei, Ricardo and Stewart, Craig and Farinha, Ana C and Lavie, Alon. COMET : A Neural Framework for MT Evaluation. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020. doi:10.18653/v1/2020.emnlp-main.213

    2020 doi

  60. [68]

    2023 , eprint=

    BLOOM: A 176B-Parameter Open-Access Multilingual Language Model , author=. 2023 , eprint=

    2023

  61. [69]

    Smith and Mike Lewis , title=

    Ofir Press and Noah A. Smith and Mike Lewis , title=. ICLR , crossref=. 2022 , cdate=

    2022

  62. [70]

    Transactions of the Association for Computational Linguistics , volume =

    Shliazhko, Oleh and Fenogenova, Alena and Tikhonova, Maria and Kozlova, Anastasia and Mikhailov, Vladislav and Shavrina, Tatiana , title =. Transactions of the Association for Computational Linguistics , volume =. 2024 , month =. doi:10.1162/tacl_a_00633 , url =

    2024 doi

  63. [71]

    OpenAI blog , volume=

    Language models are unsupervised multitask learners , author=. OpenAI blog , volume=

  64. [72]

    Few-shot Learning with Multilingual Generative Language Models

    Lin, Xi Victoria and Mihaylov, Todor and Artetxe, Mikel and Wang, Tianlu and Chen, Shuohui and Simig, Daniel and Ott, Myle and Goyal, Naman and Bhosale, Shruti and Du, Jingfei and Pasunuru, Ramakanth and Shleifer, Sam and Koura, Punit Singh and Chaudhary, Vishrav and O ' Horo,...

    2022

  65. [73]

    Advances in Neural Information Processing Systems , volume=

    Sequence to Sequence Learning with Neural Networks , author=. Advances in Neural Information Processing Systems , volume=

  66. [74]

    2014 , eprint=

    Sequence to Sequence Learning with Neural Networks , author=. 2014 , eprint=

    2014

  67. [75]

    Unsupervised Cross-lingual Representation Learning at Scale

    Conneau, Alexis and Khandelwal, Kartikay and Goyal, Naman and Chaudhary, Vishrav and Wenzek, Guillaume and Guzm \'a n, Francisco and Grave, Edouard and Ott, Myle and Zettlemoyer, Luke and Stoyanov, Veselin. Unsupervised Cross-lingual Representation Learning at Scale. Proceedin...

    2020 doi

  68. [76]

    IEEE Trans

    Pan, Sinno Jialin and Yang, Qiang , title =. IEEE Trans. on Knowl. and Data Eng. , month = oct, pages =. 2010 , issue_date =. doi:10.1109/TKDE.2009.191 , abstract =

    2010 doi

  69. [77]

    Proceedings of the 40th International Conference on Machine Learning , pages =

    The Unreasonable Effectiveness of Few-shot Learning for Machine Translation , author =. Proceedings of the 40th International Conference on Machine Learning , pages =. 2023 , editor =

    2023

  70. [78]

    Prompting P a LM for Translation: Assessing Strategies and Performance

    Vilar, David and Freitag, Markus and Cherry, Colin and Luo, Jiaming and Ratnakar, Viresh and Foster, George. Prompting P a LM for Translation: Assessing Strategies and Performance. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume ...

    2023 doi

  71. [79]

    Proceedings of the 40th Annual Meeting on Association for Computational Linguistics , pages =

    Papineni, Kishore and Roukos, Salim and Ward, Todd and Zhu, Wei-Jing , title =. Proceedings of the 40th Annual Meeting on Association for Computational Linguistics , pages =. 2002 , publisher =. doi:10.3115/1073083.1073135 , abstract =

    2002

  72. [80]

    chr F : character n-gram F -score for automatic MT evaluation

    Popovi \'c , Maja. chr F : character n-gram F -score for automatic MT evaluation. Proceedings of the Tenth Workshop on Statistical Machine Translation. 2015. doi:10.18653/v1/W15-3049

    2015 doi

  73. [81]

    chr F ++: words helping character n-grams

    Popovi \'c , Maja. chr F ++: words helping character n-grams. Proceedings of the Second Conference on Machine Translation. 2017. doi:10.18653/v1/W17-4770

    2017 doi

  74. [82]

    A Human Judgement Corpus and a Metric for A rabic MT Evaluation

    Bouamor, Houda and Alshikhabobakr, Hanan and Mohit, Behrang and Oflazer, Kemal. A Human Judgement Corpus and a Metric for A rabic MT Evaluation. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing ( EMNLP ). 2014. doi:10.3115/v1/D14-1026

    2014 doi

  75. [83]

    Computational Linguistics , volume =

    Reiter, Ehud , title =. Computational Linguistics , volume =. 2018 , month =. doi:10.1162/coli_a_00322 , url =

    2018 doi

  76. [84]

    A Call for Clarity in Reporting BLEU Scores

    Post, Matt. A Call for Clarity in Reporting BLEU Scores. Proceedings of the Third Conference on Machine Translation: Research Papers. 2018. doi:10.18653/v1/W18-6319

    2018 doi

  77. [85]

    CAM e L Tools: An Open Source Python Toolkit for A rabic Natural Language Processing

    Obeid, Ossama and Zalmout, Nasser and Khalifa, Salam and Taji, Dima and Oudah, Mai and Alhafni, Bashar and Inoue, Go and Eryani, Fadhl and Erdmann, Alexander and Habash, Nizar. CAM e L Tools: An Open Source Python Toolkit for A rabic Natural Language Processing. Proceedings of...

    2020

  78. [86]

    Cross-Lingual Empirical Evaluation of Large Language Models for A rabic Medical Tasks

    Abouzahir, Chaimae and Ma, Congbo and Habash, Nizar and Shamout, Farah E. Cross-Lingual Empirical Evaluation of Large Language Models for A rabic Medical Tasks. Proceedings of the 1st Workshop on Linguistic Analysis for Health ( H ea L ing 2026). 2026. doi:10.18653/v1/2026.hea...

    2026 doi

  79. [87]

    Viability of Machine Translation for Healthcare in Low-Resourced Languages

    Nigatu, Hellina Hailu and Mehandru, Nikita and Abadi, Negasi Haile and Gebremeskel, Blen and Alaa, Ahmed and Choudhury, Monojit. Viability of Machine Translation for Healthcare in Low-Resourced Languages. Proceedings of the 2025 Conference on Empirical Methods in Natural Langu...

    2025 doi

  80. [88]

    Findings of the WMT 2019 Biomedical Translation Shared Task: Evaluation for MEDLINE Abstracts and Biomedical Terminologies

    Bawden, Rachel and Bretonnel Cohen, Kevin and Grozea, Cristian and Jimeno Yepes, Antonio and Kittner, Madeleine and Krallinger, Martin and Mah, Nancy and Neveol, Aurelie and Neves, Mariana and Soares, Felipe and Siu, Amy and Verspoor, Karin and Vicente Navarro, Maika. Findings...

    2019

  81. [89]

    2025 , eprint=

    Pivot Language for Low-Resource Machine Translation , author=. 2025 , eprint=

    2025

  82. [90]

    How Far can 100 Samples Go? Unlocking Zero-Shot Translation with Tiny Multi-Parallel Data

    Wu, Di and Tan, Shaomu and Meng, Yan and Stap, David and Monz, Christof. How Far can 100 Samples Go? Unlocking Zero-Shot Translation with Tiny Multi-Parallel Data. Findings of the Association for Computational Linguistics: ACL 2024. 2024. doi:10.18653/v1/2024.findings-acl.896

    2024 doi

  83. [91]

    International Conference on Learning Representations , year=

    Decoupled Weight Decay Regularization , author=. International Conference on Learning Representations , year=

  84. [92]

    2020 , eprint=

    HuggingFace's Transformers: State-of-the-art Natural Language Processing , author=. 2020 , eprint=

    2020

  85. [93]

    Parameter-Efficient Transfer Learning for

    Houlsby, Neil and Giurgiu, Andrei and Jastrzebski, Stanislaw and Morrone, Bruna and De Laroussilhe, Quentin and Gesmundo, Andrea and Attariyan, Mona and Gelly, Sylvain , booktitle =. Parameter-Efficient Transfer Learning for. 2019 , editor =

    2019

  86. [94]

    von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin , license =

  87. [95]

    2022 , PAGES =

    Raiomond Doctor and Alexander Gutkin and Cibu Johny and Brian Roark and Richard Sproat , EDITOR =. 2022 , PAGES =

    2022

  88. [96]

    Oman medical journal , volume=

    Implications of language barriers for healthcare: a systematic review , author=. Oman medical journal , volume=

  89. [97]

    M oses: Open Source Toolkit for Statistical Machine Translation

    Koehn, Philipp and Hoang, Hieu and Birch, Alexandra and Callison-Burch, Chris and Federico, Marcello and Bertoldi, Nicola and Cowan, Brooke and Shen, Wade and Moran, Christine and Zens, Richard and Dyer, Chris and Bojar, Ond r ej and Constantin, Alexandra and Herbst, Evan. M o...

    2007

  90. [98]

    Aho and Jeffrey D

    Alfred V. Aho and Jeffrey D. Ullman , title =. 1972

    1972

  91. [99]

    Publications Manual , year = "1983", publisher =

    1983

  92. [100]

    Chandra and Dexter C

    Ashok K. Chandra and Dexter C. Kozen and Larry J. Stockmeyer , year = "1981", title =. doi:10.1145/322234.322243

    1981

  93. [101]

    Scalable training of

    Andrew, Galen and Gao, Jianfeng , booktitle=. Scalable training of

  94. [102]

    Dan Gusfield , title =. 1997

    1997

  95. [103]

    Tetreault , title =

    Mohammad Sadegh Rasooli and Joel R. Tetreault , title =. Computing Research Repository , volume =. 2015 , url =

    2015

  96. [104]

    A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =

    Ando, Rie Kubota and Zhang, Tong , Issn =. A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =. Journal of Machine Learning Research , Month = dec, Numpages =

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.