Rotate King to get Queen: Word Relationships as Orthogonal Transformations in Embedding Space

Kawin Ethayarajh

REVIEW 3 major objections 5 minor 26 references

Rotate King to get Queen: Word Relationships as Orthogonal Transformations in Embedding Space

T0 review · 3 major / 5 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 The paper argues that word relationships in embedding space can be represented as orthogonal or linear transformations with accuracy comparable to translation vectors, suggesting models need not encode relationships as translations.

desk verdict A novel idea about orthogonal relationship representations, but the evaluation is circular, so the accuracy claims are not established. read the letter →

arxiv 1909.00504 v2 pith:PIJZN5U6 submitted 2019-09-02 cs.CL

Kawin Ethayarajh This is my paper · ORCID

classification cs.CL

keywords wordembeddingsanalogiesorthogonaltransformationslineartranslationvectorsProcrustesembeddinggeometrydownstreamNLPmodels

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

Word relationships in embedding spaces are usually treated as translations: king minus man plus woman lands near queen. This paper argues that the same relationships can be carried by orthogonal transformations (rotations plus reflections) and by general linear maps, and that these alternatives are not just theoretical. Fitting an orthogonal matrix to approximate the mean translation vector for a relationship solves analogy-style mappings with 0.761 average accuracy, versus 0.782 for translations; a general linear map does better still, at 0.798. If the claim holds, analogy tests do not prove that embeddings store relationships as translations, and downstream models could learn relationships with purely linear components such as attention heads.

What carries the argument

The central object is a relationship-specific map $f$ that sends a source vector $x$ to a target $y$: translation $x \mapsto x + b$, orthogonal $x \mapsto R x$ with $R^\top R = I$, or linear $x \mapsto A x$. To build $R$, the paper solves the orthogonal Procrustes problem, minimizing $\| \Omega X - Y \|_F$ subject to $\Omega^\top \Omega = I$, where $X$ stacks $n$ randomly sampled word vectors and $Y$ shifts them by the mean translation vector $b$; $A$ is the ordinary least squares solution $Y X^\top (X X^\top)^{-1}$. Scoring the same transformations on analogy pairs then measures whether a simple linear map can carry the relationship.

What would settle it

Split each analogy category in half: compute the mean direction and fit the orthogonal and linear maps on the first half, then score on the second half; if accuracy falls to chance, the headline result depends on the transformations seeing the test pairs' average.

Watch

Extended reading notes

Core claim

The paper's central claim is that a word relationship can be represented as an orthogonal or linear transformation of the embedding space with accuracy comparable to, and sometimes better than, the standard translation vector. Concretely, the author takes the mean difference vector b for a relationship, samples random word vectors to form a source matrix X and a target matrix Y = X + b, and solves the orthogonal Procrustes problem for R and the ordinary least squares problem for A. On analogy categories from Mikolov et al. (2013a), the orthogonally transformed source lands on the correct target 76.1% of the time, the translation 78.2%, and the linear map 79.8%; cosine similarity to the true target is nearly equal, suggesting the result is not an artifact of a sparse candidate set. The conclusion is that parsimonious linear representations of relationships exist between simple translations and complex non-linear maps.

Load-bearing premise

The comparison assumes that letting the rotation or linear map learn the average direction of the very word pairs that will be used to score it is a fair measure of how well it represents the relationship; if that learning step leaks the answers, the reported accuracies are overstated.

Editorial extensions

If this is right

Analogy-test accuracy alone no longer distinguishes translation-based from rotation-based representations, so arithmetic analogy performance should be re-examined as a measure of embedding quality.
A single linear layer or attention head has enough capacity to carry a semantic or syntactic relationship, so attention heads found to track syntax may also track relationships such as gender.
Debiasing by subtracting a bias subspace from word vectors may miss bias encoded as a rotation or linear map, allowing gender associations to persist after projection-based debiasing.
Because orthogonal maps preserve inner products, an orthogonally transformed embedding space still allows recovery of the implicitly factorized word-context matrix, which a translated space does not.

Reading between the lines

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

A held-out evaluation that computes the mean translation vector only from training pairs would test whether orthogonal maps match translations because they genuinely represent the relationship or because the fitting step uses the same pairs that are later scored.
The same Procrustes fitting could be applied to contextualized embeddings to ask whether attention heads and feed-forward layers actually encode relationships as rotations; the paper raises this possibility but does not test it.
Fitting a random orthogonal matrix as a baseline would show how much of the near-parity is due to the flexibility of linear maps rather than to the specific structure of the embeddings.

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.

Desk Editor's Note

read the letter

I read the Rotate King paper. The idea is clean: instead of a translation vector, a word relationship can be a rotation or a general linear map. Fitting those maps with closed-form Procrustes or least squares from a random sample of words is neat, and the connection to attention heads is a nice speculative payoff. The paper is honest about not recommending the method for analogy solving. What's actually new is showing that a single rotation can approximate the gender or country-capital relationship almost as well as translation on the tested pairs. That's interesting because it gives a parsimonious alternative to translation and suggests models might learn relationships in more than one geometric form. The soft spot is not subtle. The translation vector b is computed by averaging differences over all word pairs in a category. That same b is then used to construct R and A, and the same word pairs are the test set. So the accuracy numbers measure how well R and A mimic the operation x -> x+b on random words, and how well b happens to fit the pairs it was computed from. It is not an independent test of whether rotations generalize to held-out exemplars of the relationship. The paper essentially acknowledges this when it says it is learning to approximate a translation and that it is not solving analogies in the traditional sense, but it still presents Table 1 as evidence that linear transforms are more accurate than translations. With no held-out pairs and no error bars, those differences could well be noise. The central conceptual claim survives, but the comparative accuracy claims do not. That said, the flaw is fixable. Hold out some pairs to compute b, fit R and A on random words plus that b, and evaluate on the remaining pairs. Also report variance across random samples. The paper is short and the math is correct, so a serious referee could request that and likely get a solid contribution. The citation pattern looks fine; related work is covered. I would bring this to reading group as a cautionary tale about circular evaluation, and I would cite the conceptual framing despite the empirical weakness. It deserves peer review, but with a request for a proper held-out evaluation.

Referee Report

3 major / 5 minor

Summary. The paper studies whether word relationships in embedding spaces can be represented by orthogonal or general linear transformations, rather than only by translation vectors. For each analogy category, the author computes the mean translation vector b by averaging pairwise differences over all ordered pairs in that category (Eq. 1), constructs source and target matrices X and Y = X + b from n = 2000 random vocabulary words, solves the orthogonal Procrustes problem (Eq. 2) and the ordinary least squares problem (Eq. 3) to obtain R and A, and then evaluates how accurately each transformation maps source words to target words in the same set of pairs used to define b. The reported results give average accuracies of 0.761 for orthogonal, 0.782 for translative, and 0.798 for linear transformations on GloVe, with similar trends on FastText. The paper argues that these are parsimonious representations of word relationships and discusses implications for embedding evaluation, Transformer attention heads, and debiasing.

Significance. If the empirical claim were established with an independent evaluation, the paper would provide a useful existence proof: closed-form orthogonal and linear maps can approximate translation-based analogical structure, which is relevant to understanding how downstream models might encode semantic and syntactic relationships. The paper is clearly written, uses standard closed-form solvers, and Figure 1's saturation in n and consistency across GloVe and FastText are informative. However, the current evaluation is not independent: the same translation vector both defines the transformations and is derived from the test pairs. The core novelty therefore remains unverified, and the paper's main quantitative claims require re-analysis with held-out pairs and sampling variability. The authors' explicit statement in Section 3 that they are 'ultimately learning to approximate a translation' is honest, but it also highlights that Table 1 primarily measures how well R and A mimic b on random points and how well b self-consistently fits the pairs from which it was computed.

major comments (3)

[Section 4.1, Eq. (1), Table 1] The evaluation is circular. Equation (1) computes b by averaging y - x over all word pairs in a category, and Section 4.1 states that each transformation is evaluated on 'this set of word pairs' — that is, the same pairs used to define b. Equations (2) and (3) then fit R and A to approximate the map x -> x + b on a random sample of words, so both matrices are explicitly trained to mimic b. Scoring on the pairs used to define b cannot distinguish whether the orthogonal or linear transformation represents the relationship itself or merely reproduces b's fit to its own training data. The central claim in Section 4.2 that linear transformations are more accurate than translations, and the abstract's statement that these transformations 'can be as good a representation of word relationships as translation vectors,' are therefore not supported by Table 1 as reported. Please re-run the evaluation with held-out pairs: compute b on one subset of pairs, fit R and A, and evaluate on a disjoint subset of pairs from the same relationship. Report both in-category and cross-category generalization, and clarify whether the relationship is defined by the training subset or by an external resource.
[Section 4.2, Table 1] No measure of sampling variability is reported. The three average accuracies (0.761, 0.782, 0.798) differ by at most 0.037, but the matrices R and A depend on a random sample of n = 2000 words, and the paper gives no error bars, no repeated draws, and no significance test. A single random draw could easily shift these numbers by a nontrivial amount. Please report means and standard deviations over multiple random seeds (or bootstrap over word pairs), and test whether the linear-vs-translation and orthogonal-vs-translation differences are statistically reliable. This is load-bearing for the paper's main comparative claim.
[Section 4.2, cosine-similarity column] The average cosine-similarity results are also computed on the same pairs used to define b, so they do not mitigate the circularity concern. At most, they show that the transformed source vectors are close to the target vectors on the training pairs; they do not establish that the transformations generalize to unseen exemplars of the relationship. Please report cosine similarity on held-out pairs as well, and consider whether this metric is meaningfully different from measuring how well R x approximates x + b when b was fitted to the same pairs.

minor comments (5)

[Section 4.1] Please specify the exact accuracy criterion: is a pair counted correct only if the transformed source vector's nearest vocabulary word equals the target, and is the candidate set restricted to exclude the source word and possibly other words in the same category? This matters for comparability with standard analogy evaluations such as 3CosAdd.
[Table 1] The 'Avg' row appears to be the unweighted mean over categories; since categories differ substantially in the number of pairs, please report the number of pairs per category and either weighted accuracy or both weighted and unweighted averages.
[Section 4.2, Figure 1] The paper states that accuracy plateaus for n >= 250 and similarity for n >= 500, yet the main results use n = 2000. Please explain the choice of n = 2000 and confirm that the qualitative ranking of the three methods is stable for n between 250 and 2000.
[Section 4.3] The claim that a single attention head has sufficient capacity to represent a word relationship is an extrapolation from a global linear map on static embeddings; attention heads apply learned linear maps in a context-dependent way, so the implications for Transformer architecture should be stated more cautiously.
[Throughout] There are several typos and formatting issues: 'unsupservised' should be 'unsupervised', 'V oita' should be 'Voita', and the mathematical notation (e.g., bold vectors) is inconsistent in places.

Circularity Check

1 steps flagged · score 6.0 of 10

The evaluation is circular: R and A are fit to approximate the translation vector b, which is itself computed from the same word pairs used for scoring, so the reported accuracies are partly forced.

fitted input called prediction [Section 4.1 (Task and Setup), using Eq. (1), (2), (3) from Section 3]
"we first calculate the mean translation vector b by averaging difference vectors across all word pairs, as defined in (1). b is also used to estimate matrices for the orthogonal and linear transformations (see (2) and (3)). We then create a set of word pairs for each analogy category: e.g., { (Berlin, Germany), (Paris, France), ... } for country-capital. Each type of transformation – translative, orthogonal, and linear – is evaluated by how accurately it maps source words to target words in this set of word pairs."

Eq. (1) computes b as the mean of y - x over exactly the pairs in the evaluation set S. Eqs. (2) and (3) then fit R and A to map random word vectors X to X + b, i.e., to mimic the translation defined by that same b. The task then scores the translative, orthogonal, and linear representations on the very set S that was used to compute b. The translative accuracy is an in-sample fit (b is the least-squares mean over S), and the orthogonal and linear accuracies measure how well R and A approximate b on the source words of S. Thus the reported 'accuracy' is partly forced: it is not a held-out test of whether the relationship generalizes, but a consistency check between the fitted translation and transformations trained to approximate it.

full rationale

The paper is transparent that it takes the translation vector b as given and that R and A are fit to approximate translation by b; the existence of an orthogonal or linear map that approximates a translation on a large random sample is a nontrivial empirical fact with some independent content. However, the central empirical comparison in Table 1 is not an independent benchmark: the same word pairs that define b are used for scoring, and R and A are trained to mimic that b. Consequently, high accuracy for the orthogonal and linear representations mostly reflects how well they approximate the fitted translation on the pairs from which that translation was computed. The translative baseline is itself fit to the test set by Eq. (1). No held-out relationship pairs are used, so the reported accuracies cannot support a claim of generalization to unseen exemplars of the relationship. The self-citations to Ethayarajh et al. (2019a) are motivational rather than load-bearing for the main result. Because the central evaluation reduces in part, by the paper's own equations, to fitting R and A to a b computed from the test set, the circularity score is 6 rather than a lower value.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The main free quantities are n and b. b is particularly important because it is derived from the same word pairs used for evaluation, so the transformation is not independently validated. The axioms are standard linear algebra plus the domain-specific assumption that relationships are invertible maps and that b summarizes them.

free parameters (2)

n (number of sampled words) = 2000 (main results); accuracy plateaus at 250-500
The only hyperparameter, chosen by hand. No tuning or theoretical justification beyond the plateau observation in Figure 1.
b (mean translation vector per relationship) = Computed from the evaluation word pairs (Eq. 1)
The transformation is defined relative to b, and b is fitted to the same pairs used for scoring, making the evaluation dependent on this fitted quantity.

assumptions (3)

standard math Closed-form solutions to orthogonal Procrustes and ordinary least squares are valid.
Used to compute R and A in Eqs. (2) and (3).
domain assumption Word relationships are representable as invertible transformations over embedding vectors.
Definition 1 formalizes the problem this way; it is an assumption about the nature of semantic relationships.
ad hoc to paper The mean translation vector b computed from a set of ordered pairs is a sufficient statistic for the relationship, such that a transformation fit to approximate b on random words transfers to the relationship pairs.
This is the load-bearing premise behind the evaluation in Section 4; if false, the reported accuracies do not reflect generalization.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Rotate King to get Queen: Word Relationships as Orthogonal Transformations in Embedding Space." pith.science (2026). https://pith.science/paper/PIJZN5U6

@misc{pith2026190900504,
  author       = {Pith},
  title        = {Pith review of: Rotate King to get Queen: Word Relationships as Orthogonal Transformations in Embedding Space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PIJZN5U6}},
  note         = {Machine review of arXiv:1909.00504}
}

abstract

A notable property of word embeddings is that word relationships can exist as linear substructures in the embedding space. For example, $\textit{gender}$ corresponds to $\vec{\textit{woman}} - \vec{\textit{man}}$ and $\vec{\textit{queen}} - \vec{\textit{king}}$. This, in turn, allows word analogies to be solved arithmetically: $\vec{\textit{king}} - \vec{\textit{man}} + \vec{\textit{woman}} \approx \vec{\textit{queen}}$. This property is notable because it suggests that models trained on word embeddings can easily learn such relationships as geometric translations. However, there is no evidence that models $\textit{exclusively}$ represent relationships in this manner. We document an alternative way in which downstream models might learn these relationships: orthogonal and linear transformations. For example, given a translation vector for $\textit{gender}$, we can find an orthogonal matrix $R$, representing a rotation and reflection, such that $R(\vec{\textit{king}}) \approx \vec{\textit{queen}}$ and $R(\vec{\textit{man}}) \approx \vec{\textit{woman}}$. Analogical reasoning using orthogonal transformations is almost as accurate as using vector arithmetic; using linear transformations is more accurate than both. Our findings suggest that these transformations can be as good a representation of word relationships as translation vectors.

Figures

Figures reproduced from arXiv: 1909.00504 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 15 canonical work pages

[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]

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]

Yoshua Bengio, R \'e jean Ducharme, Pascal Vincent, and Christian Jauvin. 2003. A neural probabilistic language model. Journal of Machine Learning Research, 3(Feb):1137--1155

work page 2003
[4]

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

2017
[5]

Tolga Bolukbasi, Kai-Wei Chang, James Y Zou, Venkatesh Saligrama, and Adam T Kalai. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. In Advances in Neural Information Processing Systems, pages 4349--4357

2016
[6]

Kevin Clark, Urvashi Khandelwal, Omer Levy, and Christopher D Manning. 2019. What does bert look at? an analysis of bert's attention. arXiv preprint arXiv:1906.04341

arXiv 2019
[7]

Alexis Conneau, Guillaume Lample, Marc'Aurelio Ranzato, Ludovic Denoyer, and Herv \'e J \'e gou. 2018. Word translation without parallel data. In International Conference on Learning Representations

2018
[8]

Aleksandr Drozd, Anna Gladkova, and Satoshi Matsuoka. 2016. Word embeddings, analogies, and machine learning: Beyond king-man+ woman= queen. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 3519--3530

work page 2016

Show all 26 references

[9]

Kawin Ethayarajh, David Duvenaud, and Graeme Hirst. 2019 a . https://www.aclweb.org/anthology/P19-1315 Towards understanding linear word analogies . In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 3253--3262, Florence, Italy. A...

2019
[10]

Kawin Ethayarajh, David Duvenaud, and Graeme Hirst. 2019 b . Understanding undesirable word embedding associations. In Proceedings of the 57th Conference of the Association for Computational Linguistics, pages 1696--1705

2019
[11]

Hila Gonen and Yoav Goldberg. 2019. Lipstick on a pig: Debiasing methods cover up systematic gender biases in word embeddings but do not remove them. arXiv preprint arXiv:1903.03862

2019 arXiv
[12]

Thomas K Landauer and Susan T Dumais. 1997. A solution to Plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge. Psychological Review, 104(2):211

1997
[13]

Omer Levy and Yoav Goldberg. 2014 a . Linguistic regularities in sparse and explicit word representations. In Proceedings of the Eighteenth Conference on Computational Natural Language Learning, pages 171--180

2014
[14]

Omer Levy and Yoav Goldberg. 2014 b . Neural word embedding as implicit matrix factorization. In Advances in Neural Information Processing Systems, pages 2177--2185

2014
[15]

Tomas Mikolov, Kai Chen, Greg S Corrado, and Jeff Dean. 2013 a . Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781

2013 arXiv
[16]

Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013 b . Distributed representations of words and phrases and their compositionality. In Advances in Neural Information Processing Systems, pages 3111--3119

2013
[17]

W James Murdoch, Peter J Liu, and Bin Yu. 2018. Beyond word importance: Contextual decomposition to extract interactions from LSTMs . In Proceedings of the 6th International Conference on Learning Representations (ICLR)

2018
[18]

Sungjoon Park, JinYeong Bak, and Alice Oh. 2017. Rotated word vector representations and their interpretability. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 401--411

2017
[19]

Jeffrey Pennington, Richard Socher, and Christopher Manning. 2014. GloVe : Global vectors for word representation. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1532--1543

2014
[20]

Anna Rogers, Aleksandr Drozd, and Bofang Li. 2017. The (too many) problems of analogical reasoning with word vectors. In Proceedings of the 6th Joint Conference on Lexical and Computational Semantics (* SEM 2017), pages 135--148

2017
[21]

Sascha Rothe, Sebastian Ebert, and Hinrich Sch \"u tze. 2016. Ultradense word embeddings by orthogonal transformation. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 767--777

2016
[22]

Natalie Schluter. 2018. The word analogy testing caveat. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers), pages 242--246

2018
[23]

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

1966
[24]

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in Neural Information Processing Systems

2017
[25]

Elena Voita, David Talbot, Fedor Moiseev, Rico Sennrich, and Ivan Titov. 2019. Analyzing multi-head self-attention: Specialized heads do the heavy lifting, the rest can be pruned. arXiv preprint arXiv:1905.09418

2019 arXiv
[26]

Jieyu Zhao, Tianlu Wang, Mark Yatskar, Ryan Cotterell, Vicente Ordonez, and Kai-Wei Chang. 2019. Gender bias in contextualized word embeddings. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language ...

2019

Pith tools

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

[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]

Yoshua Bengio, R \'e jean Ducharme, Pascal Vincent, and Christian Jauvin. 2003. A neural probabilistic language model. Journal of Machine Learning Research, 3(Feb):1137--1155

work page 2003

[4] [4]

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

2017

[5] [5]

Tolga Bolukbasi, Kai-Wei Chang, James Y Zou, Venkatesh Saligrama, and Adam T Kalai. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. In Advances in Neural Information Processing Systems, pages 4349--4357

2016

[6] [6]

Kevin Clark, Urvashi Khandelwal, Omer Levy, and Christopher D Manning. 2019. What does bert look at? an analysis of bert's attention. arXiv preprint arXiv:1906.04341

arXiv 2019

[7] [7]

Alexis Conneau, Guillaume Lample, Marc'Aurelio Ranzato, Ludovic Denoyer, and Herv \'e J \'e gou. 2018. Word translation without parallel data. In International Conference on Learning Representations

2018

[8] [8]

Aleksandr Drozd, Anna Gladkova, and Satoshi Matsuoka. 2016. Word embeddings, analogies, and machine learning: Beyond king-man+ woman= queen. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 3519--3530

work page 2016