Pith. sign in

REVIEW 3 major objections 5 minor 41 references

Domain adapted machine translation: What does catastrophic forgetting forget and why?

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Target vocabulary coverage predicts how much domain-adapted MT forgets

desk verdict A genuinely useful new metric and a candid empirical study, but the coverage-causality claim is shakier than the abstract suggests and Appendix B undercuts it more than the authors let on. read the letter →

arxiv 2412.17537 v1 pith:XZU7CRWV submitted 2024-12-23 cs.CL

Danielle Saunders , Steve DeNeefe This is my paper · ORCID
classification cs.CL
keywords catastrophicforgettingneuralmachinetranslationdomainadaptationvocabularycoveragemixedfine-tuningshiftForgetGenUsequalitymetrics
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

Neural machine translation models that are fine-tuned to a new domain do not forget uniformly; this paper claims they forget generic vocabulary roughly in proportion to how little of that vocabulary appears in the adaptation data. It introduces ForgetGenUse, a token-level measure showing that what is forgotten is mostly out-of-domain vocabulary, replaced by adaptation-set tokens that are often semantically related but contextually wrong, sometimes changing meaning (billion to million). Across eight domains in German-English and English-Japanese, forgetting measured by $\Delta$BLEU correlates with target-side vocabulary coverage at Kendall's $\tau=0.7$, and $\Delta$COMET at $\tau=0.9$, while common heuristics such as dataset size, segment length, likelihood, and vocabulary divergence do not. As an intervention, Minimal Mix-in adds generic sentences only until the target vocabulary is covered; it uses less than 10% of the generic data of a standard 1:1 mix-in while restoring at least 80% of the achievable forgetting mitigation on most domains and preserving in-domain quality. If this is right, practitioners could predict and prevent much of catastrophic forgetting from properties of the adaptation data alone.

What carries the argument

The central object is target vocabulary coverage, defined as the proportion of the pre-trained model's target-side subword vocabulary that appears at least once in the adaptation corpus, with source coverage measured separately. The paper's new measure ForgetGenUse counts, per reference segment, how many occurrences of a token the original model produced correctly but the adapted model no longer produces, normalized over the reference counts, so it isolates detrimental vocabulary shift rather than net quality change. Minimal Mix-in is the intervention that operationalizes the mechanism: it adds generic training sentences only until every missing target subword has been seen, thereby setting coverage near 100% while changing nothing about model architecture, adaptation, or inference. Coverage supplies the explanatory variable, ForgetGenUse supplies the vocabulary-level readout, and Minimal Mix-in converts the correlation into a controlled comparison across domains.

What would settle it

Fix one domain and dataset size while varying only target coverage, for example by deleting rare target tokens or by adding generic sentences that cover only the generic test-set vocabulary; if two such adaptation sets with different global coverage produce the same $\Delta$BLEU and $\Delta$COMET on the generic test set, then global coverage is not the driver. Alternatively, rerun the eight adaptations across several seeds and bootstrap the Kendall's $\tau$ values; if the $\tau=0.7$ and $\tau=0.9$ correlations do not survive, the paper's central claim fails.

Watch

Extended reading notes

Core claim

The paper's central discovery is that catastrophic forgetting in domain-adapted NMT is governed by the adaptation dataset's target-side vocabulary coverage: across five de-en and three en-ja domains, forgetting measured by $\Delta$BLEU correlates with target coverage at $\tau=0.7$ and $\Delta$COMET at $\tau=0.9$, while data size, segment length, generic likelihood, and vocabulary divergence do not significantly explain it. The paper also discovers that the forgetting itself is predominantly a shift away from out-of-domain vocabulary: out-of-domain tokens are forgotten at a higher rate than in-domain tokens, and the replacements are adaptation-set tokens used in inappropriate or even meaning-changing contexts, sometimes triggered by a single subword in the target vocabulary. It shows that this vocabulary shift is not captured by BLEU or COMET, since ForgetGenUseAll does not significantly correlate with either metric. The causal reading is supported by Minimal Mix-in: raising target coverage to 100% with a small generic set reduces forgetting to roughly the level of 1:1 generic mixing, with less than 10% of the generic data for most domains, and without sacrificing in-domain scores.

Load-bearing premise

The load-bearing premise is that the observed correlation between forgetting and target vocabulary coverage reflects coverage itself, rather than dataset size, the random composition of the mix-in set, or chance, since the correlations rest on eight domains with no repeated seeds and the paper's own control shows a same-size random generic set performs nearly as well.

Editorial extensions

If this is right

  • Adaptation risk becomes measurable in advance: a low target-coverage score on the adaptation data predicts larger drops in generic BLEU and COMET, so practitioners can decide whether mixing in generic data is necessary before fine-tuning.
  • Coverage-driven mixing can replace a 1:1 generic mix for most domains: using under 10% of the generic data recovers at least 80% of the achievable forgetting mitigation, with in-domain scores close to unmixed fine-tuning.
  • Quality metrics alone underestimate forgetting: a model can hold or improve BLEU/COMET while replacing correct generic tokens with wrong adaptation-set tokens, so evaluations should check token-level vocabulary shift.
  • Forgetting can fire without any in-domain trigger in the source, because a single target subword that is frequent in the adaptation set can bias the decoder, so source-side domain detection will miss some cases.
  • Equalizing target coverage to 100% makes forgetting comparable across very different domains, suggesting that the adaptation data's vocabulary, rather than the domain label itself, is what drives the loss.

Reading between the lines

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

  • The paper's Appendix B leaves implicit a sharper version of the claim: a random generic set of the same size as Minimal Mix-in reduces forgetting almost as much despite much lower global coverage, because both reach around 99% coverage of the generic test set. The operative quantity may therefore be coverage of the evaluation distribution, not of the full model vocabulary, and this distinction sho
  • If coverage is the driver, then adaptation-set curation could be inverted: instead of mixing generic data in afterward, one could select in-domain sentences whose target tokens complement the generic vocabulary, potentially preventing forgetting at the data-selection stage.
  • The limitations note that the authors see the same adaptation-set token preference in LLM-based MT in concurrent work; re-running the coverage correlations and the Minimal Mix-in intervention on a prompted or fine-tuned LLM would show whether the mechanism extends beyond the Transformer NMT setting tested here.
  • Practical use of the $\tau=0.7$ and $\tau=0.9$ correlations should wait for repeated-seed and more-language confirmation, since the statistics come from eight domain points without confidence intervals; a production rule such as 'cover 100% of the target vocabulary' could be validated by bootstrap resampling across many more domains.
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

3 major / 5 minor

Summary. The paper studies catastrophic forgetting in NMT domain adaptation. It introduces a token-level metric, ForgetGenUse, to measure vocabulary-shift forgetting, and applies it to eight domains across two language pairs (de-en and en-ja) by fine-tuning pre-trained models for a fixed 20K steps. The authors examine correlations between forgetting metrics and several domain heuristics, and report that target-side generic vocabulary coverage of the adaptation set is strongly inversely correlated with ΔBLEU (τ=0.7) and ΔCOMET (τ=0.9). They then propose Minimal Mix-in, which adds generic sentences that cover missing target vocabulary, and claim that it mitigates at least 80% of the forgetting reduction achievable by standard 1:1 random mix-in while using less than 10% of the generic data. Appendix B includes a Random #Minimal control matched for the number of generic segments, which appears to perform similarly to Minimal Mix-in despite much lower full-vocabulary coverage.

Significance. If the central claim holds, the paper offers both a diagnostic tool (vocabulary coverage) for predicting forgetting and a practical, low-cost mitigation strategy (coverage-based mix-in). The ForgetGenUse metric is a genuine contribution: it separates vocabulary shift from corpus-level quality decline and is defined independently of the outcome metrics. The paper is also transparent about its limitations and includes a control condition that partially undercuts its own causal interpretation, which is scientifically commendable. However, the correlational evidence is based on a very small number of domains and single runs, and the intervention, as analyzed, does not cleanly isolate coverage from the quantity of generic data. The strength of the contribution therefore depends on whether the authors can reconcile or sharpen the causal claim; the current manuscript is better suited as a correlational study with a plausible mechanism than as a demonstration of why forgetting occurs.

major comments (3)
  1. [§4.1 and Appendix B] The claim that Minimal Mix-in demonstrates a causal role for target-side generic vocabulary coverage is not secured. The Random #Minimal control in Appendix B, matched for the number of generic segments, reduces forgetting nearly as much as Minimal Mix-in despite much lower full-vocabulary coverage (0.65–0.70 vs 0.95–0.97 for de-en/en-ja). The reconciliation in the appendix redefines coverage relative to the generic test-set vocabulary, whereas Section 3.3 and Table 7 define coverage over the full generic training vocabulary. This metric shift means the intervention does not isolate coverage from generic-data quantity. The paper should either adopt the test-set-relative construct as the primary definition and validate it consistently, or design a comparison that varies coverage while holding mix-in size constant (e.g., random subsets of the same size with different test-set coverage). Without this, the causal 'why' in the title is not established beyond correlation.
  2. [§3.3, Table 7] The correlation analysis rests on only eight domain-level observations (five de-en, three en-ja) from single fine-tuning runs. The reported Kendall's tau values (e.g., ΔCOMET vs target coverage τ=0.9, ΔBLEU vs target coverage τ=0.7) are not accompanied by confidence intervals, and may be driven by the extreme points (Kor for de-en, BSD for en-ja). To support the phrase 'strong relationship,' the paper should report bootstrap confidence intervals, run multiple seeds, or add more domains. Without such evidence, the coverage–forgetting link remains suggestive rather than established.
  3. [§2.3] The conclusion that detrimental vocabulary shift can occur even when the source sentence has no in-domain triggers is based on a small set of manually inspected examples and fast-align-based alignments with no reported reliability measure. The paper states that these replacements 'tend to occur in very different contexts' but does not provide a systematic quantification across the full set of forgotten tokens. A more systematic analysis, or at least an inter-annotator agreement check on a larger sample, is needed to support this general claim, which is central to the paper's 'what is forgotten' contribution.
minor comments (5)
  1. [Throughout] The notation 'F orgetGenU se' appears throughout the manuscript with spacing artifacts; it should be rendered consistently as 'ForgetGenUse'.
  2. [Table 3] The column headers of Table 3 are difficult to follow; please clarify that the left-side counts are from the generic model outputs and the right-side counts are from the adapted model, with the frequency in the adaptation dataset shown in parentheses.
  3. [§3.1] The text reads 'Datasets size is recognized' and should be 'Dataset size is recognized'.
  4. [Appendix B] The term 'enja' appears where 'en-ja' is intended.
  5. [Appendix A / Reproducibility] The paper does not mention whether code or trained models will be released; given the experimental nature, a statement on data/code availability would help reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the coverage correlations and Minimal Mix-in are genuine empirical results, though Appendix B's random control raises a confound that is a correctness issue, not a circular reduction.

full rationale

The paper is an empirical study whose quantitative claims do not reduce to their inputs by construction. The new ForgetGenUse metric (Equation 1) is defined from reference translations and pre-/post-adaptation model outputs, independently of the vocabulary-coverage heuristic; coverage is defined in Section 3.3 as “the proportion of the generic subword vocabulary that appears at all in the preprocessed segments,” again independently of the forgetting metrics. The Table 7 correlations are computed across eight domains without fitting any parameter to the forgetting values, so no “fitted input called prediction” pattern is present. Minimal Mix-in is a genuine intervention: generic lines are selected to cover missing target tokens, and the claim that it mitigates at least 80% of Random 1:1’s mitigation is a measured outcome, not a definitional identity. The potentially load-bearing weakness is Appendix B’s Random #Minimal control, which “mitigates forgetting similarly on average to Minimal Mix-in, despite their apparently different coverage statistics”; the authors reconcile this by switching from full-vocabulary coverage to generic-test-set coverage. That reconciliation is post hoc and weakens the causal interpretation, but it is an alternative-explanation concern rather than circularity: no equation or fitted parameter makes the outcome equal to the input. The self-citations present (e.g., Saunders 2022 as a survey and for the open definition of domain) are not load-bearing, and the paper imports no uniqueness theorem or ansatz from prior author work. Under the circularity-only mandate, the appropriate finding is no significant circularity, with the Appendix B confound flagged as a correctness risk rather than a circular step.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper's central claims rest on a small number of domains, a specific definition of coverage, and a fixed adaptation protocol. The new measure ForgetGenUse is a definition rather than a free parameter, but the analysis choices around it (subsampling target sizes, LASER cutoff) are hand-chosen. No new physical or conceptual entities are introduced.

free parameters (3)
  • LASER filtering cutoff = 0.8
    Selected by inspection in Section 3.2 for Sub-ssf and Kor-filtered datasets. Affects the quality-filtered subsampling experiments, not the central correlation directly.
  • Subsampling target token count = Approximately Kor/BSD token counts
    In Section 3.1, de-en datasets are subsampled to roughly the token count of Kor, and en-ja to BSD. This controls for size but the target is arbitrary and hand-chosen.
  • Fine-tuning step count = 20K steps
    All adaptation experiments use 20K steps, chosen to trigger forgetting comparable to prior work. This is a fixed design choice, not fitted to the outcome, but it affects all results and may interact with coverage.
assumptions (5)
  • domain assumption WMT News/General concatenated test sets represent 'generic' language
    Used as the generic test sets for measuring forgetting. If these sets are not generic, the forgetting measurements are biased.
  • domain assumption Subword-level unigram coverage is a meaningful measure of vocabulary coverage
    The central correlate is based on BPE subword coverage. The paper does not validate against word-level or n-gram coverage.
  • domain assumption fast_align symmetrized alignments correctly identify token replacements
    Section 2.3 uses fast_align to determine which tokens changed and what replaced them. Alignment errors could misattribute forgetting.
  • standard math Kendall's tau with n=5 or n=3 can support the reported correlations
    The paper treats tau values as significant with p<0.05 despite the very small number of domains, without correcting for multiple comparisons.
  • ad hoc to paper The 20K-step fine-tuning protocol isolates data effects
    All experiments use the same adaptation length. The relationship between coverage and forgetting may depend on this specific training regime.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Domain adapted machine translation: What does catastrophic forgetting forget and why?." pith.science (2026). https://pith.science/paper/XZU7CRWV

@misc{pith2026241217537,
  author       = {Pith},
  title        = {Pith review of: Domain adapted machine translation: What does catastrophic forgetting forget and why?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XZU7CRWV}},
  note         = {Machine review of arXiv:2412.17537}
}
read the original abstract

Neural Machine Translation (NMT) models can be specialized by domain adaptation, often involving fine-tuning on a dataset of interest. This process risks catastrophic forgetting: rapid loss of generic translation quality. Forgetting has been widely observed, with many mitigation methods proposed. However, the causes of forgetting and the relationship between forgetting and adaptation data are under-explored. This paper takes a novel approach to understanding catastrophic forgetting during NMT adaptation by investigating the impact of the data. We provide a first investigation of what is forgotten, and why. We examine the relationship between forgetting and the in-domain data, and show that the amount and type of forgetting is linked to that data's target vocabulary coverage. Our findings pave the way toward better informed NMT domain adaptation.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 20 canonical work pages

  1. [1]

    Roee Aharoni and Yoav Goldberg. 2020. https://doi.org/10.18653/v1/2020.acl-main.692 Unsupervised domain clusters in pretrained language models . In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7747--7763, Online. Association for Computational Linguistics

    doi:10.18653/v1/2020.acl-main.692 2020

  2. [2]

    Antonio Valerio Miceli Barone, Barry Haddow, Ulrich Germann, and Rico Sennrich. 2017. https://doi.org/10.18653/v1/D17-1156 Regularization techniques for fine-tuning in Neural Machine Translation . In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 1489--1494

    work page doi:10.18653/v1/d17-1156 2017

  3. [3]

    Costa-juss \`a , Fethi Bougares, and Olivier Galibert

    Lo \" c Barrault, Magdalena Biesialska, Marta R. Costa-juss \`a , Fethi Bougares, and Olivier Galibert. 2020. https://aclanthology.org/2020.wmt-1.2 Findings of the first shared task on lifelong learning machine translation . In Proceedings of the Fifth Conference on Machine Translation, pages 56--64, Online. Association for Computational Linguistics

    work page 2020

  4. [4]

    Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2017. https://doi.org/10.1162/tacl_a_00051 Enriching word vectors with subword information . Transactions of the Association for Computational Linguistics, 5:135--146

    doi:10.1162/tacl_a_00051 2017

  5. [5]

    Mauro Cettolo, Christian Girardi, and Marcello Federico. 2012. https://aclanthology.org/2012.eamt-1.60 WIT 3: Web inventory of transcribed and translated talks . In Proceedings of the 16th Annual Conference of the European Association for Machine Translation, pages 261--268, Trento, Italy. European Association for Machine Translation

    work page 2012

  6. [6]

    Chenhui Chu, Raj Dabre, and Sadao Kurohashi. 2017. https://doi.org/10.18653/v1/P17-2061 An empirical comparison of domain adaptation methods for neural machine translation . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 385--391, Vancouver, Canada. Association for Computational L...

    work page doi:10.18653/v1/p17-2061 2017

  7. [7]

    Chris Dyer, Victor Chahuneau, and Noah A. Smith. 2013. https://www.aclweb.org/anthology/N13-1073 A simple, fast, and effective reparameterization of IBM model 2 . In Proceedings of the 2013 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies , pages 644--648, Atlanta, Georgia. Association...

    work page 2013

  8. [8]

    Shuhao Gu and Yang Feng. 2020. https://doi.org/10.18653/v1/2020.coling-main.381 Investigating catastrophic forgetting during continual training for neural machine translation . In Proceedings of the 28th International Conference on Computational Linguistics, pages 4315--4326, Barcelona, Spain (Online). International Committee on Computational Linguistics

    doi:10.18653/v1/2020.coling-main.381 2020

Show all 41 references
  1. [9]

    Rejwanul Haque, Yasmin Moslem, and Andy Way. 2020. https://aclanthology.org/2020.icon-adapmt.4 Terminology-aware sentence mining for NMT domain adaptation: ADAPT ' s submission to the adap- MT 2020 E nglish-to- H indi AI translation shared task . In Proceedings of the 17th Int...

    2020

  2. [10]

    Eva Hasler, Tobias Domhan, Jonay Trenous, Ke Tran, Bill Byrne, and Felix Hieber. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.666 Improving the quality trade-off for neural machine translation multi-domain adaptation . In Proceedings of the 2021 Conference on Empirical Me...

    2021 doi

  3. [11]

    Huda Khayrallah and Philipp Koehn. 2018. https://doi.org/10.18653/v1/W18-2709 On the impact of various types of noise on neural machine translation . In Proceedings of the 2nd Workshop on Neural Machine Translation and Generation, pages 74--83, Melbourne, Australia. Associatio...

    2018 doi

  4. [12]

    Tom Kocmi, Eleftherios Avramidis, Rachel Bawden, Ond r ej Bojar, Anton Dvorkovich, Christian Federmann, Mark Fishel, Markus Freitag, Thamme Gowda, Roman Grundkiewicz, Barry Haddow, Philipp Koehn, Benjamin Marie, Christof Monz, Makoto Morishita, Kenton Murray, Makoto Nagata, To...

    2023 doi

  5. [13]

    Tom Kocmi, Rachel Bawden, Ond r ej Bojar, Anton Dvorkovich, Christian Federmann, Mark Fishel, Thamme Gowda, Yvette Graham, Roman Grundkiewicz, Barry Haddow, Rebecca Knowles, Philipp Koehn, Christof Monz, Makoto Morishita, Masaaki Nagata, Toshiaki Nakazawa, Michal Nov \'a k, Ma...

    2022

  6. [14]

    J. Lin. 1991. https://doi.org/10.1109/18.61115 Divergence measures based on the shannon entropy . IEEE Transactions on Information Theory, 37(1):145--151

    1991 doi

  7. [15]

    Jinghui Lu, Maeve Henchion, and Brian Mac Namee. 2020. https://aclanthology.org/2020.lrec-1.832 Diverging divergences: Examining variants of J ensen S hannon divergence for corpus comparison tasks . In Proceedings of the Twelfth Language Resources and Evaluation Conference, pa...

    2020

  8. [16]

    Minh-Thang Luong and Christopher D Manning. 2015. Stanford Neural Machine Translation systems for spoken language domains . In Proceedings of the International Workshop on Spoken Language Translation, pages 76--79

    2015

  9. [17]

    Michael McCloskey and Neal J Cohen. 1989. Catastrophic interference in connectionist networks: The sequential learning problem. In Psychology of learning and motivation, volume 24, pages 109--165. Elsevier

    1989

  10. [18]

    Makoto Morishita, Katsuki Chousa, Jun Suzuki, and Masaaki Nagata. 2022. https://aclanthology.org/2022.lrec-1.721 JP ara C rawl v3.0: A large-scale E nglish- J apanese parallel corpus . In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 6704--6...

    2022

  11. [19]

    Graham Neubig. 2011. The Kyoto free translation task. http://www.phontron.com/kftt

    2011

  12. [20]

    Jianhui Pang, Fanghua Ye, Longyue Wang, Dian Yu, Derek F Wong, Shuming Shi, and Zhaopeng Tu. 2024. Salute the classic: R evisiting challenges of machine translation in the age of large language models. arXiv preprint arXiv:2401.08350

    2024 arXiv

  13. [21]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. https://doi.org/10.3115/1073083.1073135 B leu: a method for automatic evaluation of machine translation . In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311--3...

    2002

  14. [22]

    Minh Quang Pham, Josep Maria Crego, Fran c ois Yvon, and Jean Senellart. 2020. https://aclanthology.org/2020.wmt-1.72 A study of residual adapters for multi-domain neural machine translation . In Proceedings of the Fifth Conference on Machine Translation, pages 617--628, Onlin...

    2020

  15. [23]

    Matt Post. 2018. https://doi.org/10.18653/v1/W18-6319 A call for clarity in reporting BLEU scores . In Proceedings of the Third Conference on Machine Translation: Research Papers, pages 186--191, Brussels, Belgium. Association for Computational Linguistics

    2018 doi

  16. [24]

    Ricardo Rei, Craig Stewart, Ana C Farinha, and Alon Lavie. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.213 COMET : A neural framework for MT evaluation . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 2685--2702, ...

    2020 doi

  17. [25]

    Mat \= ss Rikters, Ryokan Ri, Tong Li, and Toshiaki Nakazawa. 2019. https://doi.org/10.18653/v1/D19-5204 Designing the business conversation corpus . In Proceedings of the 6th Workshop on Asian Translation, pages 54--61, Hong Kong, China. Association for Computational Linguistics

    2019 doi

  18. [26]

    Danielle Saunders. 2022. Domain adaptation and multi-domain adaptation for neural machine translation: A survey. Journal of Artificial Intelligence Research, 75:351--424

    2022

  19. [27]

    Danielle Saunders and Bill Byrne. 2020. https://aclanthology.org/2020.wmt-1.94 Addressing exposure bias with document minimum risk training: C ambridge at the WMT 20 biomedical translation task . In Proceedings of the Fifth Conference on Machine Translation, pages 862--869, On...

    2020

  20. [28]

    Konstantin Savenkov and Michel Lopez. 2022. https://aclanthology.org/2022.amta-upg.4 The state of the machine translation 2022 . In Proceedings of the 15th Biennial Conference of the Association for Machine Translation in the Americas (Volume 2: Users and Providers Track and G...

    2022

  21. [29]

    Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. https://doi.org/10.18653/v1/P16-1162 Neural machine translation of rare words with subword units . In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages ...

    2016 doi

  22. [30]

    Chenze Shao and Yang Feng. 2022. https://doi.org/10.18653/v1/2022.acl-long.143 Overcoming catastrophic forgetting beyond continual learning: Balanced training for neural machine translation . In Proceedings of the 60th Annual Meeting of the Association for Computational Lingui...

    2022 doi

  23. [31]

    Amr Sharaf, Hany Hassan, and Hal Daum \'e III. 2020. https://doi.org/10.18653/v1/2020.ngt-1.5 Meta-learning for few-shot NMT adaptation . In Proceedings of the Fourth Workshop on Neural Generation and Translation, pages 43--53, Online. Association for Computational Linguistics

    2020 doi

  24. [32]

    Brian Thompson, Jeremy Gwinnup, Huda Khayrallah, Kevin Duh, and Philipp Koehn. 2019. https://doi.org/10.18653/v1/N19-1209 Overcoming catastrophic forgetting during domain adaptation of neural machine translation . In Proceedings of the 2019 Conference of the North A merican Ch...

    2019 doi

  25. [33]

    Marlies van der Wees, Arianna Bisazza, Wouter Weerkamp, and Christof Monz. 2015. https://doi.org/10.3115/v1/P15-2092 What ' s in a domain? A nalyzing genre and topic differences in statistical machine translation . In Proceedings of the 53rd Annual Meeting of the Association f...

    2015 doi

  26. [34]

    Dusan Varis and Ond r ej Bojar. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.650 Sequence length is a domain: Length-based overfitting in transformer models . In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 8246--8257, Onli...

    2021 doi

  27. [35]

    Ashish Vaswani, Samy Bengio, Eugene Brevdo, Francois Chollet, Aidan Gomez, Stephan Gouws, Llion Jones, ukasz Kaiser, Nal Kalchbrenner, Niki Parmar, Ryan Sepassi, Noam Shazeer, and Jakob Uszkoreit. 2018. https://aclanthology.org/W18-1819 T ensor2 T ensor for neural machine tran...

    2018

  28. [36]

    Yu Wan, Baosong Yang, Derek Fai Wong, Lidia Sam Chao, Liang Yao, Haibo Zhang, and Boxing Chen. 2022. https://doi.org/10.1162/coli_a_00435 Challenges of neural machine translation for short texts . Computational Linguistics, 48(2):321--342

    2022 doi

  29. [37]

    Jitao Xu, Josep Crego, and Jean Senellart. 2019. https://aclanthology.org/2019.iwslt-1.27 Lexical micro-adaptation for neural machine translation . In Proceedings of the 16th International Conference on Spoken Language Translation, Hong Kong. Association for Computational Linguistics

    2019

  30. [38]

    Xuan Zhang, Navid Rajabi, Kevin Duh, and Philipp Koehn. 2023. https://doi.org/10.18653/v1/2023.wmt-1.43 Machine translation with large language models: Prompting, few-shot learning, and fine-tuning with QL o RA . In Proceedings of the Eighth Conference on Machine Translation, ...

    2023 doi

  31. [39]

    Xuan Zhang, Pamela Shapiro, Gaurav Kumar, Paul McNamee, Marine Carpuat, and Kevin Duh. 2019. https://doi.org/10.18653/v1/N19-1189 Curriculum learning for domain adaptation in neural machine translation . In Proceedings of the 2019 Conference of the North A merican Chapter of t...

    2019 doi

  32. [40]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  33. [41]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.