Pith. sign in

REVIEW 2 major objections 5 minor 44 references

The Devil Is in the Word Alignment Details: On Translation-Based Cross-Lingual Transfer for Token Classification Tasks

T0 review · 2 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Optimized word alignment matches or beats marker-based label projection, and an ensemble of the two translation directions outperforms it.

desk verdict A solid empirical study showing that with proper tuning, word alignment label projection matches marker-based methods, and the ETT ensemble is a simple, robust gain; the main caveat is the unquantified discontinuous-span problem in verb-final languages. read the letter →

arxiv 2505.10507 v2 pith:V6DSV7JZ submitted 2025-05-15 cs.CL

Benedikt Ebing , Goran Glavaš This is my paper · ORCID
classification cs.CL
keywords cross-lingualtransfertokenclassificationlabelprojectionwordalignmenttranslate-traintranslate-testensembleslotlabeling
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

To do translation-based cross-lingual transfer on token-level tasks, a system has to move labels from the original sentence onto its translation; this paper studies the word-alignment route for that label projection. It claims that earlier comparisons against tag-based (marker) projection were unfair to word aligners, because big gains come from three details: projecting whole labeled spans instead of token-by-token labels, filtering alignments that look incomplete, and choosing how the translated sentence is pre-tokenized. With those details set well, word aligners match or beat marker-based projection on both translate-train and translate-test over 29 languages, and the paper's ensemble of the two translation directions (ETT), which averages the class logits returned on each token, outperforms the marker-based projection and shrinks the effect of the low-level choices. A sympathetic reader should care because word-alignment projection needs no special fine-tuning of the translation model, so the result makes translation-based transfer cheaper and its behavior more predictable.

What carries the argument

The argument rests on four mechanisms. The first is span-based projection: rather than moving token labels one by one along alignments, it takes a labeled span in the source sentence, maps the BIO start tag of the first source token onto the first aligned target token, maps the continuation tag across all intervening aligned target tokens, and thereby enforces continuous well-formed spans in the translation. The second is a set of filters that discard or shrink projections whose alignments look incomplete: source tokens with no target link, target spans that are discontinuous, or span counts and types that do not match, with a dedicated restricted-target rule for single-token spans in translate-test. The third is pre-tokenization of the translated sentence with a language-specific tokenizer rather than plain whitespace. The fourth is ETT, which projects the translate-test class logits back into the target sentence and averages them with translate-train logits token-wise. These mechanisms carry the argument because they convert noisy alignments into conservative, well-formed spans and then let the two translation directions correct each other.

What would settle it

Take a language with verb-final word order, or any pair where reordering splits spans, annotate gold target spans, and measure how often a single source span's aligned target tokens form two or more disconnected chunks. Run word-aligner-based translate-test and ETT only on that subset: if the F1 gap versus marker-based projection is large on discontinuous examples while small on continuous ones, the continuity assumption, not alignment quality, is what carries the result.

Watch

Extended reading notes

Core claim

Contrary to the conclusions of recent marker-based projection papers, optimized word-alignment projection delivers transfer performance at least on par with inserting and preserving tags around labeled spans, and occasionally better: on named-entity recognition it slightly exceeds the marker methods in translate-train, and on slot labeling it trails by less than a point. Translate-test, which translates the target sentence to English for prediction, is the sensitive setting: a language-specific pretokenizer adds up to 12.6 points on the NER benchmark, and a filtering rule that restricts single-token source spans to single target tokens recovers about 9 points over applying no filter. Translate-train, in contrast, is largely indifferent to these choices. The ETT ensemble, averaging the class logits of translate-train and translate-test for every target token, improves average F1 by about 3.1 points over translate-train and 1.5 points over translate-test on the two benchmarks, and it cuts the loss from using whitespace instead of language-specific tokenization from about 9.9 points to 1.4.

Load-bearing premise

The load-bearing premise is that one labeled span in the original sentence always corresponds to one continuous labeled span in the translation; the authors concede this fails for some German and Dutch examples, where sentence-final verbs split a slot into disjoint pieces, so frequent discontinuous projections would break the projection and the ensemble.

Editorial extensions

If this is right

  • Translate-test should receive the tuning effort: language-specific pre-tokenization and restricted-target filtering buy up to 12.6 points on the NER benchmark, while translate-train choices barely move the score.
  • ETT lifts average F1 over translate-train by about 3.1 points and over translate-test by about 1.5 points, and reduces the gap between good and bad pre-tokenization choices from roughly 9.9 to 1.4 points.
  • Across the reported experiments, the comparison holds with different translation engines, different word aligners, and different downstream encoders, so the result is not tied to a single model choice.
  • Word-aligner projection avoids the tag-preservation fine-tuning or constrained decoding that marker-based methods require, so the matched performance makes the cheaper option the default.

Reading between the lines

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

  • The logit-averaging trick in ETT should generalize to ensembling more than two projections, say several aligners or tokenizers at translate-test time, and would likely compress design-sensitivity further at the cost of extra translation compute; the paper only demonstrates the two-component version.
  • The continuity assumption sets a ceiling: for verb-final languages or free-word-order pairs where discontinuous target spans are common, ETT would need a projection that can emit multiple target spans per source span; the paper notes the phenomenon for German and Dutch but does not build that extension.
  • Because translate-train is insensitive to the filtering choices, a practical recipe for low-resource languages is to fix one reasonable translate-train configuration and spend all tuning budget on translate-test, then combine with ETT.
  • Benchmark reports of translation-based transfer should state label-projection details; otherwise, differences between word-aligner and marker-based systems are not interpretable, which is the source of the contradiction with earlier results.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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. This paper revisits word-aligner (WA) based label projection for translation-based cross-lingual transfer (XLT) on token classification tasks. The authors systematically investigate three low-level design choices: the span-mapping algorithm, filtering strategies for discarding incomplete alignments, and pre-tokenization of the translated sentence. Using MasakhaNER 2.0 and xSID (29 languages, English as source), they show that these choices substantially affect translate-train (T-Train) and especially translate-test (T-Test) performance. With configurations selected on validation data, they report that WA-based projection matches or outperforms the marker-based methods Easy and Codec. They further propose ETT, an ensemble that averages the class logits of a T-Train and a T-Test model after projecting the latter back to the original target sentence, and show that ETT improves over both components and reduces sensitivity to the pre-tokenization choice. Additional experiments vary the base encoder/decoder model, the word aligner, and the MT system.

Significance. The paper is a valuable, carefully executed empirical study. It directly challenges the recent claim that marker-based label projection is inherently superior to WA-based projection, and it provides a clear recipe for configuring WA-based pipelines. The evaluation is broad and transparent: two benchmarks, 29 languages, per-language tables, multiple base models, aligners, and MT systems, and the code and data are publicly released. The proposed ETT ensemble is simple and effective, and the finding that it reduces sensitivity to design choices is practically important. The main weaknesses are an overgeneralized robustness claim and an unquantified limitation regarding discontinuous projected spans; both are addressable in revision. In its current form, the evidence supports the average 'WA matches or surpasses' claim, but the robustness claim requires either additional experiments or rephrasing.

major comments (2)
  1. [Abstract; Section 4.2, Table 3; Section 6] The claim that ETT 'reduces sensitivity to low-level WA design choices' (abstract, contribution 3) and 'mitigates the observed performance variations making suboptimal filtering strategies less influential' (Limitations) is not fully supported by the experiments. Table 3 demonstrates robustness only for the pre-tokenization choice (WS-TOK vs LS-TOK) in T-Test. No experiment shows ETT with different filtering strategies (e.g., NO-FILT versus RSTR-TGT, which differ by 9.2 points for T-Test in Table 1) or with different span-mapping algorithms. Please either add an ETT ablation over the filtering strategies from Table 1, or qualify the claims so that 'low-level design choices' refers specifically to the pre-tokenization (and, where shown, MT model) dimensions.
  2. [Section 2.1.1; Section 6; Table 15] The span-mapping algorithm assumes that every labeled source span projects to a single continuous target span, and the Limitations state that this fails for some German and Dutch xSID examples because sentence-final verbs split a slot into discontinuous spans. The paper does not quantify how often this occurs, nor does it provide an ablation that repairs discontinuous projections. This matters because Table 15 shows that WA-based T-Train lags Codec by 2.6 F1 on German and 4.4 F1 on Dutch, well above the 0.9 average lag, a pattern consistent with systematic mislabeling of affected spans. I ask the authors to report the frequency of discontinuous spans and, ideally, to evaluate a repair strategy (e.g., splitting the projected span) so that the impact on the 'matches or surpasses' and robustness claims for verb-final languages can be assessed.
minor comments (5)
  1. [Table 4] The row headers 'AccAlginnof t' and 'Awesomenof t' are formatting errors and should read 'AccAlign (no ft)' and 'Awesome (no ft)'.
  2. [Appendix G] 'per-tokenization experiments' should be 'pre-tokenization experiments'.
  3. [Tables 14 and 15] The section header 'Ensembling-Translate-Train' does not match the 'Ensemble-Train-Test' (ETT) terminology used in the main text and should be renamed.
  4. [Section 3 and Section 4.2] The WA configuration is selected on validation data while the marker-based baselines use published defaults; the discussion should explicitly acknowledge this asymmetry when interpreting the 'matches or surpasses' claim, since it is favorable to WA.
  5. [Results throughout] Differences are reported as means over three seeds without significance testing; for Table 2, where several differences are 0.3-0.6 F1 points with standard deviations of 0.4-0.7, a paired test or a comment on effect size would strengthen the comparative claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: all central claims are empirical comparisons against external benchmarks and published baselines, with no fitted parameter or self-citation chain determining the outcome.

full rationale

This paper contains no derivation chain whose output is fixed by its inputs. The central claims—that WA design choices substantially affect translation-based XLT, that an optimized WA-based label projection matches or surpasses marker-based methods, and that the proposed ETT ensemble improves performance and robustness—are established by direct empirical comparison on MasakhaNER 2.0 and xSID against published baselines (Easy, Codec) and external components (AccAlign, Awesome, NLLB, Google Translate). The low-level choices (span mapping, filtering strategies, pre-tokenization) are evaluated as ablations on validation data, and the main results are reported on test portions; selecting a configuration on validation is standard model selection, not a fitted-parameter-as-prediction scheme. The self-citations to Ebing and Glavaš (2024) support only auxiliary practices such as training on original plus translated data and the observation that T-Test is more sensitive to translation quality; these are not load-bearing because the comparative claims are computed from the paper's own experiments rather than imported from those citations. The Limitations admission that the continuity assumption fails for some German and Dutch xSID examples is an acknowledged robustness caveat and a potential correctness risk, but it does not make any prediction equivalent to an input by construction. No circular step meets the evidentiary bar of quoting a specific equation or fitted parameter that reduces to the claimed result, so the appropriate score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claims are empirical, not derivational. The only free elements are categorical design choices tuned on validation and standard hyperparameters; there are no invented entities or fitted constants embedded in a derivation.

free parameters (3)
  • Filtering strategy choices = T-Train: COMP-INS+COMP-TGT+COMP-SRC; T-Test: RSTR-TGT
    Selected from validation results in Tables 1, 6 and 7; categorical design choices that change T-Test by up to 9.2 F1.
  • Pre-tokenization choices = T-Train: whitespace (WS-TOK), Chinese uses LS-TOK; T-Test: language-specific (LS-TOK)
    Selected on validation in Tables 12 and 13; LS-TOK improves T-Test by up to 12.6%.
  • Training hyperparameters = lr=1e-5, 10 epochs, batch 32, beam size 5
    Standard values from prior work; not central to the comparative claims.
assumptions (4)
  • domain assumption Labeled spans remain continuous under translation.
    Section 2.1.1 span mapping builds a single continuous target span; Limitations state this fails for some German and Dutch examples.
  • domain assumption Word alignments are an adequate proxy for token correspondence.
    The entire label projection pipeline relies on alignments from AccAlign or Awesome.
  • domain assumption Validation performance is representative for selecting design choices.
    The authors select filters and tokenizers on validation and apply them to test; no distribution shift analysis is provided.
  • domain assumption Translationese in translated evaluation data does not overturn the conclusions.
    Acknowledged in Limitations with reference to Artetxe et al. (2020); may slightly overestimate true transfer performance.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Devil Is in the Word Alignment Details: On Translation-Based Cross-Lingual Transfer for Token Classification Tasks." pith.science (2026). https://pith.science/paper/V6DSV7JZ

@misc{pith2026250510507,
  author       = {Pith},
  title        = {Pith review of: The Devil Is in the Word Alignment Details: On Translation-Based Cross-Lingual Transfer for Token Classification Tasks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V6DSV7JZ}},
  note         = {Machine review of arXiv:2505.10507}
}
read the original abstract

Translation-based strategies for cross-lingual transfer XLT such as translate-train -- training on noisy target language data translated from the source language -- and translate-test -- evaluating on noisy source language data translated from the target language -- are competitive XLT baselines. In XLT for token classification tasks, however, these strategies include label projection, the challenging step of mapping the labels from each token in the original sentence to its counterpart(s) in the translation. Although word aligners (WAs) are commonly used for label projection, the low-level design decisions for applying them to translation-based XLT have not been systematically investigated. Moreover, recent marker-based methods, which project labeled spans by inserting tags around them before (or after) translation, claim to outperform WAs in label projection for XLT. In this work, we revisit WAs for label projection, systematically investigating the effects of low-level design decisions on token-level XLT: (i) the algorithm for projecting labels between (multi-)token spans, (ii) filtering strategies to reduce the number of noisily mapped labels, and (iii) the pre-tokenization of the translated sentences. We find that all of these substantially impact translation-based XLT performance and show that, with optimized choices, XLT with WA offers performance at least comparable to that of marker-based methods. We then introduce a new projection strategy that ensembles translate-train and translate-test predictions and demonstrate that it substantially outperforms the marker-based projection. Crucially, we show that our proposed ensembling also reduces sensitivity to low-level WA design choices, resulting in more robust XLT for token classification tasks.

Figures

Figures reproduced from arXiv: 2505.10507 by the authors.

Figure 1
Figure 1. Schematic overview of our pipeline for label projection using word aligners. From top to bottom, we [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Transfer performance with WA for language [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Results for translation-based XLT with LLM2Vec (L) vs. DeBERTa (D), relative to zero-shot XLT performance with XLM-R (X). Masakha xSID Avg Translate-Train AccAlign 67.1±1.2 82.7±0.8 75.0±1.0 AccAlginnoft 66.7±1.1 82.9±0.5 74.9±1.0 Awesomenoft 64.4±1.3 79.8±0.8 72.7±1.2 Translate-Test AccAlign 72.3±0.5 80.2±0.3 76.3±0.4 AccAlginnoft 70.4±0.5 79.6±0.3 75.0±0.4 Awesomenoft 65.1±0.4 74.8±0.3 70.0±0.4 [PITH_FULL_IMAGE:f… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 3 canonical work pages

  1. [1]

    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 INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bamba Dione, Andiswa Bukula, Rooweither Mabuya, Bonaventure F

    David Adelani, Graham Neubig, Sebastian Ruder, Shruti Rijhwani, Michael Beukman, Chester Palen-Michel, Constantine Lignos, Jesujoba Alabi, Shamsuddeen Muhammad, Peter Nabende, Cheikh M. Bamba Dione, Andiswa Bukula, Rooweither Mabuya, Bonaventure F. P. Dossou, Blessing Sibanda, Happy Buzaaba, Jonathan Mukiibi, Godson Kalipe, Derguene Mbaye, Amelia Taylor, ...

    2022

  4. [4]

    Divyanshu Aggarwal, Vivek Gupta, and Anoop Kunchukuttan. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.755 I ndic XNLI : Evaluating multilingual inference for I ndian languages . In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 10994--11006, Abu Dhabi, United Arab Emirates. Association for Computational ...

    doi:10.18653/v1/2022.emnlp-main.755 2022

  5. [5]

    Kabir Ahuja, Harshita Diddee, Rishav Hada, Millicent Ochieng, Krithika Ramesh, Prachi Jain, Akshay Nambi, Tanuja Ganu, Sameer Segal, Mohamed Ahmed, Kalika Bali, and Sunayana Sitaram. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.258 MEGA : Multilingual evaluation of generative AI . In Proceedings of the 2023 Conference on Empirical Methods in Natural ...

    doi:10.18653/v1/2023.emnlp-main.258 2023

  6. [6]

    Mikel Artetxe, Vedanuj Goswami, Shruti Bhosale, Angela Fan, and Luke Zettlemoyer. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.399 Revisiting machine translation for cross-lingual classification . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6489--6499, Singapore. Association for Computational Linguistics

    doi:10.18653/v1/2023.emnlp-main.399 2023

  7. [7]

    Mikel Artetxe, Gorka Labaka, and Eneko Agirre. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.618 Translation artifacts in cross-lingual transfer learning . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 7674--7684, Online. Association for Computational Linguistics

    doi:10.18653/v1/2020.emnlp-main.618 2020

  8. [8]

    Parishad BehnamGhader, Vaibhav Adlakha, Marius Mosbach, Dzmitry Bahdanau, Nicolas Chapados, and Siva Reddy. 2024. Llm2vec: Large language models are secretly powerful text encoders. arXiv preprint arXiv:2404.05961

    arXiv 2024

Show all 44 references
  1. [9]

    Yang Chen, Chao Jiang, Alan Ritter, and Wei Xu. 2023. https://doi.org/10.18653/v1/2023.findings-acl.357 Frustratingly easy label projection for cross-lingual transfer . In Findings of the Association for Computational Linguistics: ACL 2023, pages 5775--5796, Toronto, Canada. A...

    2023 doi

  2. [10]

    Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzm \'a n, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. https://doi.org/10.18653/v1/2020.acl-main.747 Unsupervised cross-lingual representation learning ...

    2020 doi

  3. [11]

    Alice Coucke, Alaa Saade, Adrien Ball, Théodore Bluche, Alexandre Caulier, David Leroy, Clément Doumouro, Thibault Gisselbrecht, Francesco Caltagirone, Thibaut Lavril, Maël Primet, and Joseph Dureau. 2018. https://arxiv.org/abs/1805.10190 Snips voice platform: an embedded spok...

    2018 arXiv

  4. [12]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://doi.org/10.18653/v1/N19-1423 BERT : Pre-training of deep bidirectional transformers for language understanding . In Proceedings of the 2019 Conference of the North A merican Chapter of the Associat...

    2019 doi

  5. [13]

    Zi-Yi Dou and Graham Neubig. 2021. https://doi.org/10.18653/v1/2021.eacl-main.181 Word alignment by fine-tuning embeddings on parallel corpora . In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 2...

    2021 doi

  6. [14]

    David Duki \'c and Jan Snajder. 2024. https://doi.org/10.18653/v1/2024.findings-acl.843 Looking right is sometimes right: Investigating the capabilities of decoder-only LLM s for sequence labeling . In Findings of the Association for Computational Linguistics ACL 2024, pages 1...

    2024 doi

  7. [15]

    Chris Dyer, Victor Chahuneau, and Noah A. Smith. 2013. https://aclanthology.org/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: Hum...

    2013

  8. [16]

    Benedikt Ebing and Goran Glava s . 2024. https://doi.org/10.18653/v1/2024.naacl-long.298 To translate or not to translate: A systematic investigation of translation-based cross-lingual transfer to low-resource languages . In Proceedings of the 2024 Conference of the North Amer...

    2024 doi

  9. [17]

    Abteen Ebrahimi, Manuel Mager, Arturo Oncevay, Vishrav Chaudhary, Luis Chiruzzo, Angela Fan, John Ortega, Ricardo Ramos, Annette Rios, Ivan Vladimir Meza Ruiz, Gustavo Gim \'e nez-Lugo, Elisabeth Mager, Graham Neubig, Alexis Palmer, Rolando Coto-Solano, Thang Vu, and Katharina...

    2022 doi

  10. [18]

    Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Naveen Arivazhagan, and Wei Wang. 2022. https://doi.org/10.18653/v1/2022.acl-long.62 Language-agnostic BERT sentence embedding . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long...

    2022 doi

  11. [19]

    Iker Garc \' a-Ferrero, Rodrigo Agerri, and German Rigau. 2022. https://doi.org/10.18653/v1/2022.findings-emnlp.478 Model and data transfer for cross-lingual sequence labelling in zero-resource settings . In Findings of the Association for Computational Linguistics: EMNLP 2022...

    2022 doi

  12. [20]

    Iker Garc \' a-Ferrero, Rodrigo Agerri, and German Rigau. 2023. https://doi.org/10.18653/v1/2023.findings-emnlp.1015 T -projection: High quality annotation projection for sequence labeling tasks . In Findings of the Association for Computational Linguistics: EMNLP 2023, pages ...

    2023 doi

  13. [21]

    Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, and et al. 2024. https://arxiv.org/abs/2407.21783 The llama 3 herd of models . Preprint, arXiv:2407.21783

    2024 arXiv

  14. [22]

    Pengcheng He, Jianfeng Gao, and Weizhu Chen. 2023. https://arxiv.org/abs/2111.09543 Debertav3: Improving deberta using electra-style pre-training with gradient-disentangled embedding sharing . Preprint, arXiv:2111.09543

    2023 arXiv

  15. [23]

    Junjie Hu, Sebastian Ruder, Aditya Siddhant, Graham Neubig, Orhan Firat, and Melvin Johnson. 2020. https://proceedings.mlr.press/v119/hu20b.html XTREME : A massively multilingual multi-task benchmark for evaluating cross-lingual generalisation . In Proceedings of the 37th Inte...

    2020

  16. [24]

    Binyuan Hui, Jian Yang, Zeyu Cui, Jiaxi Yang, Dayiheng Liu, Lei Zhang, Tianyu Liu, Jiajun Zhang, Bowen Yu, Keming Lu, Kai Dang, Yang Fan, Yichang Zhang, An Yang, Rui Men, Fei Huang, Bo Zheng, Yibo Miao, Shanghaoran Quan, Yunlong Feng, Xingzhang Ren, Xuancheng Ren, Jingren Zhou...

    2024 arXiv

  17. [25]

    Masoud Jalili Sabet, Philipp Dufter, Fran c ois Yvon, and Hinrich Sch \"u tze. 2020. https://doi.org/10.18653/v1/2020.findings-emnlp.147 S im A lign: High quality word alignments without parallel training data using static and contextualized embeddings . In Findings of the Ass...

    2020 doi

  18. [26]

    Anne Lauscher, Vinit Ravishankar, Ivan Vuli \'c , and Goran Glava s . 2020. https://doi.org/10.18653/v1/2020.emnlp-main.363 From zero to hero: O n the limitations of zero-shot language transfer with multilingual T ransformers . In Proceedings of the 2020 Conference on Empirica...

    2020 doi

  19. [27]

    Duong Minh Le, Yang Chen, Alan Ritter, and Wei Xu. 2024. https://arxiv.org/abs/2402.03131 Constrained decoding for cross-lingual label projection . Preprint, arXiv:2402.03131

    2024 arXiv

  20. [28]

    Franz Josef Och and Hermann Ney. 2003. https://doi.org/10.1162/089120103321337421 A systematic comparison of various statistical alignment models . Computational Linguistics, 29(1):19--51

    2003 doi

  21. [29]

    Jaehoon Oh, Jongwoo Ko, and Se-Young Yun. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.452 Synergy with translation artifacts for training and inference in multilingual tasks . In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, page...

    2022 doi

  22. [30]

    OpenAI, Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, and et al. 2024. https://arxiv.org/abs/2303.08774 Gpt-4 technical report . Preprint, arXiv:2303.08774

    2024 arXiv

  23. [31]

    Tanmay Parekh, I-Hung Hsu, Kuan-Hao Huang, Kai-Wei Chang, and Nanyun Peng. 2024. https://doi.org/10.18653/v1/2024.naacl-long.321 Contextual label projection for cross-lingual structured prediction . In Proceedings of the 2024 Conference of the North American Chapter of the Ass...

    2024 doi

  24. [32]

    Evgeniia Razumovskaia, Ivan Vuli \'c , and Anna Korhonen. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.369 Transfer-free data-efficient multilingual slot labeling . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6041--6055...

    2023 doi

  25. [33]

    Sebastian Ruder, Noah Constant, Jan Botha, Aditya Siddhant, Orhan Firat, Jinlan Fu, Pengfei Liu, Junjie Hu, Dan Garrette, Graham Neubig, and Melvin Johnson. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.802 XTREME - R : Towards more challenging and nuanced multilingual eva...

    2021 doi

  26. [34]

    Fabian David Schmidt, Ivan Vuli \'c , and Goran Glava s . 2022. https://doi.org/10.18653/v1/2022.emnlp-main.736 Don ' t stop fine-tuning: On training regimes for few-shot cross-lingual transfer with multilingual language models . In Proceedings of the 2022 Conference on Empiri...

    2022 doi

  27. [35]

    Sebastian Schuster, Sonal Gupta, Rushin Shah, and Mike Lewis. 2019. https://doi.org/10.18653/v1/N19-1380 Cross-lingual transfer learning for multilingual task oriented dialog . In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computat...

    2019 doi

  28. [36]

    Gemma Team, Morgane Riviere, Shreya Pathak, Pier Giuseppe Sessa, Cassidy Hardin, Surya Bhupatiraju, and et al. 2024. https://arxiv.org/abs/2408.00118 Gemma 2: Improving open language models at a practical size . Preprint, arXiv:2408.00118

    2024 arXiv

  29. [37]

    NLLB Team, Marta R. Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula, Loic Barrault, Gabriel Mejia Gonzalez, Prang...

    2022 arXiv

  30. [38]

    Tjong Kim Sang and Fien De Meulder

    Erik F. Tjong Kim Sang and Fien De Meulder. 2003. https://aclanthology.org/W03-0419 Introduction to the C o NLL -2003 shared task: Language-independent named entity recognition . In Proceedings of the Seventh Conference on Natural Language Learning at HLT - NAACL 2003 , pages 142--147

    2003

  31. [39]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, and et al. 2023. https://arxiv.org/abs/2307.09288 Llama 2: Open foundation and fine-tuned chat models . Preprint, arXiv:2307.09288

    2023 arXiv

  32. [40]

    Rob van der Goot, Ibrahim Sharaf, Aizhan Imankulova, Ahmet \"U st \"u n, Marija Stepanovi \'c , Alan Ramponi, Siti Oryza Khairunnisa, Mamoru Komachi, and Barbara Plank. 2021. https://doi.org/10.18653/v1/2021.naacl-main.197 From masked language modeling to translation: Non- E n...

    2021 doi

  33. [41]

    Liang Wang, Nan Yang, Xiaolong Huang, Linjun Yang, Rangan Majumder, and Furu Wei. 2024. https://doi.org/10.18653/v1/2024.acl-long.642 Improving text embeddings with large language models . In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguisti...

    2024 doi

  34. [42]

    Weikang Wang, Guanhua Chen, Hanqing Wang, Yue Han, and Yun Chen. 2022. https://doi.org/10.18653/v1/2022.findings-emnlp.215 Multilingual sentence transformer as a multilingual word aligner . In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 2952--2...

    2022 doi

  35. [43]

    Zihan Wang, Stephen Mayhew, Dan Roth, et al. 2019. Cross-lingual ability of multilingual bert: An empirical study. arXiv preprint arXiv:1912.07840

    2019 arXiv

  36. [44]

    Shijie Wu and Mark Dredze. 2019. https://doi.org/10.18653/v1/D19-1077 Beto, bentz, becas: The surprising cross-lingual effectiveness of BERT . In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference o...

    2019 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.