Pith. sign in

REVIEW 3 major objections 4 minor 41 references

Duality Regularization for Unsupervised Bilingual Lexicon Induction

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Training both directions of a bilingual lexicon mapping as a primal–dual pair, with a cycle-consistency loss, improves accuracy and stability and yields the best published results without parallel data.

desk verdict A plausible cycle-consistency regularizer that improves stability, but the 'best published' claim rests on an unexplained semi-supervised variant and needs a major fix before it can be trusted. read the letter →

arxiv 1909.01013 v2 pith:RAPNVB5A submitted 2019-09-03 cs.CL

Xuefeng Bai , Yue Zhang , Hailong Cao , Tiejun Zhao This is my paper · ORCID
classification cs.CL
keywords bilinguallexiconinductionunsupervisedwordtranslationcycleconsistencyadversarialtrainingembeddingsdualityback-translationcross-lingualmapping
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

The paper tries to establish that bilingual lexicon induction—mapping one language's word vectors into another's—should not be solved as two independent one-way tasks. It shows that the forward and backward mappings can be treated as primal and dual problems and trained jointly, with a cycle-consistency loss forcing the round trip $x \to F(x) \to G(F(x))$ to return near $x$. On six language pairs and two benchmarks, this simple addition improves precision at 1 and makes training markedly more stable across random initializations. If true, it means unsupervised bilingual dictionaries can be built without parallel data and, when paired with iterative refinement, can match or beat supervised and semi-supervised methods.

What carries the argument

The load-bearing object is the cycle-consistency loss, $\ell_{\text{cycle}}$, which penalizes disagreement between the forward mapping and the inverse of the backward mapping by measuring average cosine similarity between original embeddings and round-tripped embeddings, $x$ versus $G(F(x))$ and $y$ versus $F(G(y))$. It converts the two adversarial min-max games into one joint objective, so the two directions cannot drift into contradictory solutions. The paper also modifies the unsupervised model-selection criterion to a weighted average of forward and backward similarity scores, which stabilizes checkpoint selection during training.

What would settle it

Reproduce the paper's best configuration using only monolingual embeddings, tracing exactly how the GeoMMsemi component is initialized; if the top Table 4 scores require a supervised warm start or a seed dictionary, the claim of a fully unsupervised state of the art is false.

Watch

Extended reading notes

Core claim

The central claim is that adding a cycle-consistency regularizer to the adversarial alignment baseline produces better and more consistent bilingual lexicons. The paper defines two linear mappings, $F: X\to Y$ and $G:Y\to X$, and adds to the adversarial loss the terms $\ell_{\text{cycle}}(F,G,X)=\frac{1}{n}\sum_i \Delta(x_i, G(F(x_i)))$ and the symmetric term $\ell_{\text{cycle}}(F,G,Y)$, with $\Delta$ set to average cosine similarity. Training both directions jointly lowers back-translation inconsistency from roughly 67–76% to 64–69% on MUSE and raises precision at 1 on both MUSE and Vecmap, with especially large gains on directions where the one-way baseline collapses. Combined with Procrustes or GeoMMsemi refinement, the fully unsupervised model obtains the best reported Vecmap scores on most language pairs.

Load-bearing premise

The strongest results rest on the assumption that the component called GeoMMsemi is being run without any seed translation pairs, but the paper calls it 'weak supervision' and never states how that supervision is eliminated in the unsupervised comparison.

Editorial extensions

If this is right

  • Cycle regularization cuts the inconsistency rate between the two translation directions on all five MUSE pairs tested, from roughly 67–76% down to 64–69%.
  • Average precision at 1 over 10 runs improves on nearly every direction, with the largest gains where the one-way baseline fails completely, such as DE-EN on Vecmap rising from 0.0 to 33.4.
  • The method is less sensitive to initialization, so a single run is more likely to land near the average accuracy rather than at a collapsed local optimum.
  • Combining the regularized model with iterative Procrustes refinement beats the Sinkhorn-BT and Adv-C-Procrustes baselines on most Vecmap language pairs.
  • The fully unsupervised model with GeoMMsemi refinement matches or exceeds supervised and semi-supervised baselines on all tested Vecmap pairs except EN-FI.

Reading between the lines

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

  • A natural extension, not tested in the paper, is to apply the same forward-backward cycle constraint to other directional alignment tasks such as cross-lingual sentence retrieval or unsupervised machine translation, where single-direction models are also known to drift.
  • Because the cycle term is measured by cosine similarity rather than exact identity, it could be satisfied by mappings that preserve similarity without recovering exact translations; this suggests the gains may shrink on typologically distant language pairs, a hypothesis that could be tested directly.
  • If the GeoMMsemi component in the best configuration is not genuinely seed-free, then the paper's contribution reduces to improving the stability of the adversarial baseline rather than establishing a new fully unsupervised state of the art; the released code would settle this.
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

3 major / 4 minor

Summary. The paper addresses unsupervised bilingual lexicon induction (UBLI) by jointly training a forward mapping F and a backward mapping G with an adversarial objective plus cycle-consistency losses (Eqs. 2-3), and a bidirectional model-selection criterion (Eq. 4). It compares with the adversarial model of Conneau et al. (2018) on MUSE and Vecmap, reporting best and average P@1 over 10 runs, and presents two variants of the proposed method: Ours-Procrustes and Ours-GeoMMsemi. The latter is claimed to be fully unsupervised and to achieve the best published results on the Vecmap benchmark.

Significance. The proposed regularization is simple, intuitive, and targets a real failure mode of adversarial UBLI, namely directional inconsistency and collapse. If the reported effects are reproducible, the method could be a useful drop-in improvement over the Conneau et al. (2018) baseline. The paper's strengths include a clear description of the objective, explicit reporting of best and average over 10 runs, and a quantitative inconsistency analysis (Table 2). The contribution is empirical rather than theoretical, and the main weaknesses are that the headline 'best published results' claim rests on an underspecified variant, and the empirical evidence lacks significance testing and released code.

major comments (3)
  1. [Section 4.3, Table 4] The claim that Ours-GeoMMsemi is 'fully unsupervised' is not supported. GeoMMsemi is defined in Section 4.3 as 'iterative GeoMM with weak supervision' and is listed under Supervised Methods, yet no text explains how the weak supervision is removed in the Ours-GeoMMsemi row of Table 4. This is load-bearing because Ours-Procrustes, the variant whose construction is fully described, is below Unsup-SL on every language pair (e.g., EN-DE 46.58/42.50 vs 48.22/44.09; EN-FI 28.08/26.48 vs 32.95/33.45). The 'best published results' claim therefore stands or falls on the Ours-GeoMMsemi row. The near-equality of Ours-GeoMMsemi and supervised GeoMMsemi on EN-IT (50.00/42.67 in both rows) makes it especially important to state exactly what supervision, if any, the GeoMMsemi refinement consumes.
  2. [Section 3.2, Eqs. (2)-(3)] The sign convention of the cycle term is ambiguous. The paper defines the discrepancy criterion Δ as 'the average cosine similarity', and Eq. (3) presents a total objective to be minimized. Minimizing an average cosine similarity would push reconstructed embeddings away from the original embeddings, contradicting the stated goal of encouraging closed loops. The authors should state explicitly whether they minimize the negative cosine similarity or use a transformed dissimilarity; without this clarification, the method cannot be reproduced from the text.
  3. [Section 4.1, Tables 1 and 4] The abstract's claim that the method 'significantly outperforms competitive baselines' is not supported by statistical evidence. The paper reports best and average accuracies over 10 runs but gives no standard deviations, per-run distributions, or significance tests. Given the known high variance of adversarial UBLI (the Adv-C baseline collapses to 0.0 on several Vecmap directions), average differences could be driven by a small number of runs. Please add variance measures or significance tests, or weaken the wording accordingly.
minor comments (4)
  1. [Section 1] The paper states 'We release our code at xxx'; no working code link is provided. Given the dependence of the results on the precise training procedure, a public implementation is important for verification.
  2. [Section 3.3, Eq. (4)] The hyperparameter λ is introduced for model selection and said to work well at 0.5, but no sensitivity analysis is reported. Please provide evidence that the results are robust to the choice of λ.
  3. [Table 2] The inconsistency rates directly reflect the training objective in Eq. (2), so lower rates are partly a mechanical consequence of the regularizer. It would be more informative to show whether the reduction in inconsistency correlates with P@1 improvements, rather than presenting the rates as independent evidence.
  4. [Section 4.3] The concurrent adversarial autoencoder method of Mohiuddin and Joty (2019) is cited but not included in Table 4; given the 'best published results' claim, the authors should explain why this concurrent method is not compared.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core contribution is an empirical regularizer tested against external P@1 benchmarks, and the flagged GeoMMsemi supervision issue is a correctness concern, not a circular derivation.

full rationale

The paper makes no claimed first-principles derivation; its contribution is an empirical cycle-consistency regularizer added to the adversarial UBLI baseline of Conneau et al. (2018). The headline results are P@1 scores on the MUSE and Vecmap benchmarks, which are external test sets with gold translation dictionaries, so the primary evaluation is not constructed from the training objective. Table 2's inconsistency rates do measure the cycle-consistency objective, but those rates are not the evidence for the best-published claim; the external P@1 comparisons are. There are no load-bearing self-citations by the present authors, and the cited cycle-consistency and dual-learning works are independent prior art explicitly acknowledged. The only notable manuscript-level concern, flagged here for completeness, is that Section 4.3 calls Ours-GeoMMsemi 'fully unsupervised' after defining GeoMMsemi as 'iterative GeoMM with weak supervision,' and the paper does not explain how weak supervision is removed; this affects the fairness of the state-of-the-art comparison and the unpublished 'code at xxx' further limits reproducibility. However, this is a supervision-labeling and evidence-quality issue, not a case where a prediction is equivalent to its input by construction or where a fitted parameter is renamed as a prediction. The derivation chain is therefore not circular.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The only hand-chosen parameter is λ; the core reliance is on the linear mapping assumption and the unsupervised selection criterion inherited from Conneau et al. 2018. The paper introduces no new entities.

free parameters (1)
  • lambda (λ) = 0.5
    Weight in the model selection score Sa (Eq. 4), set to 0.5 by hand and reported to generally work; not tuned per language pair, so its effect on results is not analyzed.
assumptions (3)
  • domain assumption Word embedding spaces of two languages can be related by a linear map F.
    Inherited from Conneau et al. 2018; the entire method and evaluation assume linear alignability, which is known to fail for distant languages such as English-Esperanto (low results in Table 1).
  • domain assumption The unsupervised model selection score Sa correlates with translation accuracy.
    Section 3.3 follows Conneau et al. 2018; the paper extends it to the weighted forward-backward average but provides no validation that Sa selects better models than random or than the original criterion.
  • ad hoc to paper Average cosine similarity is an appropriate cycle discrepancy.
    Section 3.2 sets Δ as average cosine similarity; the paper notes in Table 3 that this causes near-synonym rather than exact back-translations, showing the criterion is a weak proxy for translation identity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Duality Regularization for Unsupervised Bilingual Lexicon Induction." pith.science (2026). https://pith.science/paper/RAPNVB5A

@misc{pith2026190901013,
  author       = {Pith},
  title        = {Pith review of: Duality Regularization for Unsupervised Bilingual Lexicon Induction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RAPNVB5A}},
  note         = {Machine review of arXiv:1909.01013}
}
read the original abstract

Unsupervised bilingual lexicon induction naturally exhibits duality, which results from symmetry in back-translation. For example, EN-IT and IT-EN induction can be mutually primal and dual problems. Current state-of-the-art methods, however, consider the two tasks independently. In this paper, we propose to train primal and dual models jointly, using regularizers to encourage consistency in back translation cycles. Experiments across 6 language pairs show that the proposed method significantly outperforms competitive baselines, obtaining the best-published results on a standard benchmark.

Figures

Figures reproduced from arXiv: 1909.01013 by the authors.

Figure 1
Figure 1. (a) Inconsistency between primal model F and the dual model G. (b) An ideal scenario. Italian, the primal model maps the word “three” to the Italian word “tre”, but the dual model maps “tre” to “two” instead of “three”. We propose to address this issue by exploiting duality, encouraging forward and backward mappings to form a closed loop (Figure 1b). In particular, we extend the model of Conneau et al. (2018) by usi… view at source ↗
Figure 2
Figure 2. The proposed framework. (a) X → F(X) → G(F(X)) → X; (b) Y → G(Y ) → F(G(Y )) → Y . idea can be expressed as the minmax game minFmaxDy `adv(F, Dy, X, Y ), where `adv(F, Dy, X, Y ) = 1 m Xm j=1 logPDy (src = 1|yj ) + 1 n Xn i=1 logPDy (src = 0|F(xi)). (1) PDy (src|yj ) is a model probability from Dy to distinguish whether word embedding yj is coming from the target language (src = 1) or the primal mapping F (src = 0).… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 31 canonical work pages

  1. [1]

    URL: " 'urlintro :=

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

  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]

    Hanan Aldarmaki, Mahesh Mohan, and Mona Diab. 2018. http://aclweb.org/anthology/Q18-1014 Unsupervised word mapping using structural similarities in monolingual embeddings . Transactions of the Association for Computational Linguistics, 6:185--196

    work page 2018

  4. [4]

    David Alvarez-Melis and Tommi Jaakkola. 2018. http://aclweb.org/anthology/D18-1214 Gromov-wasserstein alignment of word embedding spaces . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 1881--1890. Association for Computational Linguistics

    work page 2018

  5. [5]

    Mart \'i n Arjovsky, Soumith Chintala, and L \'e on Bottou. 2017. Wasserstein gan. arXiv preprint arXiv:1701.07875

    arXiv 2017

  6. [6]

    Mikel Artetxe, Gorka Labaka, and Eneko Agirre. 2017. https://doi.org/10.18653/v1/P17-1042 Learning bilingual word embeddings with (almost) no bilingual data . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 451--462. Association for Computational Linguistics

    doi:10.18653/v1/p17-1042 2017

  7. [7]

    Mikel Artetxe, Gorka Labaka, and Eneko Agirre. 2018 a . http://aclweb.org/anthology/P18-1073 A robust self-learning method for fully unsupervised cross-lingual mappings of word embeddings . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 789--798. Association for Computational Linguistics

    work page 2018

  8. [8]

    Mikel Artetxe, Gorka Labaka, and Eneko Agirre. 2018 b . http://aclweb.org/anthology/D18-1399 Unsupervised statistical machine translation . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 3632--3642. Association for Computational Linguistics

    work page 2018

Show all 41 references
  1. [9]

    Mikel Artetxe, Gorka Labaka, Eneko Agirre, and Kyunghyun Cho. 2018 c . https://openreview.net/forum?id=Sy2ogebAW Unsupervised neural machine translation . In International Conference on Learning Representations

    2018

  2. [10]

    A. P. Sarath Chandar, Stanislas Lauly, Hugo Larochelle, Mitesh M. Khapra, Balaraman Ravindran, Vikas C. Raykar, and Amrita Saha. 2014. An autoencoder approach to learning bilingual word representations. In NIPS

    2014

  3. [11]

    Alexis Conneau, Guillaume Lample, Marc'Aurelio Ranzato, Ludovic Denoyer, and Herv \'e J \'e gou. 2018. https://openreview.net/forum?id=H196sainb Word translation without parallel data . In International Conference on Learning Representations

    2018

  4. [12]

    Marco Cuturi. 2013. http://papers.nips.cc/paper/4927-sinkhorn-distances-lightspeed-computation-of-optimal-transport.pdf Sinkhorn distances: Lightspeed computation of optimal transport . In C. J. C. Burges, L. Bottou, M. Welling, Z. Ghahramani, and K. Q. Weinberger, editors, Ad...

    2013

  5. [13]

    Georgiana Dinu, Angeliki Lazaridou, and Marco Baroni. 2015. Improving zero-shot learning by mitigating the hubness problem. In Proceedings of the 3th International Conference on Learning Representations (ICLR)

    2015

  6. [14]

    Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C

    Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville, and Yoshua Bengio. 2014. Generative adversarial nets. In NIPS

    2014

  7. [15]

    Jiang Guo, Wanxiang Che, David Yarowsky, Haifeng Wang, and Ting Liu. 2015. http://www.aclweb.org/anthology/P15-1119 Cross-lingual dependency parsing based on distributed representations . In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistic...

    2015

  8. [16]

    Di He, Yingce Xia, Tao Qin, Liwei Wang, Nenghai Yu, Tie-Yan Liu, and Wei-Ying Ma. 2016. http://papers.nips.cc/paper/6469-dual-learning-for-machine-translation.pdf Dual learning for machine translation . In D. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, and R. Garnett, editor...

    2016

  9. [17]

    Yedid Hoshen and Lior Wolf. 2018. http://aclweb.org/anthology/D18-1043 Non-adversarial unsupervised word translation . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 469--478. Association for Computational Linguistics

    2018

  10. [18]

    Pratik Jawanpuria, Arjun Balgovind, Anoop Kunchukuttan, and Bamdev Mishra. 2018. http://arxiv.org/abs/1808.08773 Learning multilingual word embeddings in latent metric space: A geometric approach . CoRR, abs/1808.08773

    2018 arXiv

  11. [19]

    Yova Kementchedjhieva, Sebastian Ruder, Ryan Cotterell, and Anders S gaard. 2018. http://aclweb.org/anthology/K18-1021 Generalizing procrustes analysis for better bilingual dictionary induction . In Proceedings of the 22nd Conference on Computational Natural Language Learning,...

    2018

  12. [20]

    Taeksoo Kim, Moonsu Cha, Hyunsoo Kim, Jung Kwon Lee, and Jiwon Kim. 2017. Learning to discover cross-domain relations with generative adversarial networks. In ICML

    2017

  13. [21]

    Guillaume Lample, Alexis Conneau, Ludovic Denoyer, and Marc'Aurelio Ranzato. 2018 a . https://openreview.net/forum?id=rkYTTf-AZ Unsupervised machine translation using monolingual corpora only . In International Conference on Learning Representations

    2018

  14. [22]

    Guillaume Lample, Myle Ott, Alexis Conneau, Ludovic Denoyer, and Marc'Aurelio Ranzato. 2018 b . http://aclweb.org/anthology/D18-1549 Phrase-based & neural unsupervised machine translation . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processi...

    2018

  15. [23]

    Robert Litschko, Goran Glava s , Simone Paolo Ponzetto, and Ivan Vuli\' c . 2018. https://doi.org/10.1145/3209978.3210157 Unsupervised cross-lingual information retrieval using monolingual data only . In The 41st International ACM SIGIR Conference on Research &\#38; Developmen...

    2018

  16. [24]

    Stephen Mayhew, Chen-Tse Tsai, and Dan Roth. 2017. https://doi.org/10.18653/v1/D17-1269 Cheap translation for cross-lingual named entity recognition . In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2536--2545. Association for C...

    2017 doi

  17. [25]

    Antonio Valerio Miceli Barone. 2016. https://doi.org/10.18653/v1/W16-1614 Towards cross-lingual distributed representations without parallel text trained with adversarial autoencoders . In Proceedings of the 1st Workshop on Representation Learning for NLP, pages 121--126. Asso...

    2016 doi

  18. [26]

    Le, and Ilya Sutskever

    Tomas Mikolov, Quoc V. Le, and Ilya Sutskever. 2013. http://arxiv.org/abs/1309.4168 Exploiting similarities among languages for machine translation . CoRR, abs/1309.4168

    2013 arXiv

  19. [27]

    Tasnim Mohiuddin and Shafiq R. Joty. 2019. http://arxiv.org/abs/1904.04116 Revisiting adversarial autoencoder for unsupervised word translation with cycle consistency and improved training . CoRR, abs/1904.04116

    2019 arXiv

  20. [28]

    Tanmoy Mukherjee, Makoto Yamada, and Timothy Hospedales. 2018. http://aclweb.org/anthology/D18-1063 Learning unsupervised word translations without adversaries . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 627--632. Associat...

    2018

  21. [29]

    Sch \"o nemann

    Peter H. Sch \"o nemann. 1966. https://doi.org/10.1007/BF02289451 A generalized solution of the orthogonal procrustes problem . Psychometrika, 31(1):1--10

    1966 doi

  22. [30]

    Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. https://doi.org/10.18653/v1/P16-1009 Improving neural machine translation models with monolingual data . In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ...

    2016 doi

  23. [31]

    Ivan Vuli \'c and Marie-Francine Moens. 2013. Cross-lingual semantic similarity of words as the similarity of their semantic word responses. In Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Tec...

    2013

  24. [32]

    Ivan Vuli\' c and Marie-Francine Moens. 2015. https://doi.org/10.1145/2766462.2767752 Monolingual and cross-lingual information retrieval models based on (bilingual) word embeddings . In Proceedings of the 38th International ACM SIGIR Conference on Research and Development in ...

    2015

  25. [33]

    Yijun Wang, Yingce Xia, Lu Zhao, Jiang Bian, Tao Qin, Guiquan Liu, and Tie-Yan Liu. 2018. Dual transfer learning for neural machine translation with marginal distribution regularization. In AAAI

    2018

  26. [34]

    Yingce Xia, Jiang Bian, Tao Qin, Nenghai Yu, and Tie-Yan Liu. 2017. https://doi.org/10.24963/ijcai.2017/434 Dual inference for machine learning . In Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, IJCAI-17 , pages 3112--3118

    2017 doi

  27. [35]

    Smith, and Jaime Carbonell

    Jiateng Xie, Zhilin Yang, Graham Neubig, Noah A. Smith, and Jaime Carbonell. 2018. https://arxiv.org/abs/1808.09861 Neural cross-lingual named entity recognition with minimal resources . In Conference on Empirical Methods in Natural Language Processing (EMNLP), Brussels, Belgium

    2018 arXiv

  28. [36]

    Ruochen Xu, Yiming Yang, Naoki Otani, and Yuexin Wu. 2018. http://aclweb.org/anthology/D18-1268 Unsupervised cross-lingual transfer of word embedding spaces . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 2465--2474. Associati...

    2018

  29. [37]

    Zhen Yang, Wei Chen, Feng Wang, and Bo Xu. 2018. http://aclweb.org/anthology/P18-1005 Unsupervised neural machine translation with weight sharing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 46--55....

    2018

  30. [38]

    Meng Zhang, Yang Liu, Huanbo Luan, and Maosong Sun. 2017 a . https://doi.org/10.18653/v1/P17-1179 Adversarial training for unsupervised bilingual lexicon induction . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pape...

    2017 doi

  31. [39]

    Meng Zhang, Yang Liu, Huanbo Luan, and Maosong Sun. 2017 b . https://doi.org/10.18653/v1/D17-1207 Earth mover's distance minimization for unsupervised bilingual lexicon induction . In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages...

    2017 doi

  32. [40]

    a henb \

    Tinghui Zhou, Philipp Kr \"a henb \"u hl, Mathieu Aubry, Qi-Xing Huang, and Alexei A. Efros. 2016. Learning dense correspondence via 3d-guided cycle consistency. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 117--126

    2016

  33. [41]

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. 2017. Unpaired image-to-image translation using cycle-consistent adversarial networks. In 2017 IEEE International Conference on Computer Vision (ICCV), pages 2242--2251

    2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.