Pith. sign in

REVIEW 4 major objections 4 minor 44 references

Combining Discrete Wavelet and Cosine Transforms for Efficient Sentence Embedding

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper proposes a parameter-free combination of discrete wavelet and cosine transforms that encodes sentences into fixed-size vectors of the word-embedding dimension, matching or exceeding larger baselines on several NLP tasks.

desk verdict A plausible, non-parameterized DWT-DCT sentence embedding that needs a fixed configuration and variance reporting before its headline gains over DCT-only can be trusted. read the letter →

arxiv 2508.00420 v1 pith:ZEHBHLMY submitted 2025-08-01 cs.CL

Rana Salama , Abdou Youssef , Mona Diab This is my paper
classification cs.CL
keywords sentenceembeddingdiscretewavelettransformcosinewordcompressionparameter-freespectralmethodsnon-parametricnaturallanguageprocessing
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 claims that discrete wavelet transforms (DWT) and discrete cosine transforms (DCT) can be applied together to turn variable-length sentences into fixed-size vectors without losing task performance. If true, this gives a simple, parameter-free alternative to methods that either average word vectors, losing word order, or concatenate many DCT coefficients, producing very long vectors such as 1800 dimensions for six coefficients on 300-dimensional words. The proposal applies DWT along each word vector to compress it, then DCT across the words in the sentence to encode word order, ending with a vector the size of one word embedding. Experiments on sentiment, paraphrase, and question-type tasks show results comparable to or better than averaging and DCT-only baselines, often at a quarter or half the dimension. The paper also presents evidence that DWT coefficients alone can compress word embeddings to 50–75% of their original size while preserving semantic similarity.

What carries the argument

The central object is the conjoint DWT-DCT transform, which first applies a discrete wavelet transform row-wise to the sentence matrix (each row is a word vector), splitting every word into approximation and detail coefficient vectors at one or more levels, and then applies a discrete cosine transform column-wise across the words of the sentence; the final sentence vector is formed by concatenating one or two DCT coefficients from each coefficient matrix, with a skipping trick that keeps the total dimension equal to the word-embedding dimension $d$ instead of $Kd$. This two-stage pipeline is what lets the method compress both the intra-word structure (via DWT) and the inter-word order (via DCT) into a fixed-size vector.

What would settle it

Permute the coordinate order of every word embedding with one shared random permutation and rerun the reported experiments. If downstream accuracy stays close to the unpermuted case, the wavelet stage is not exploiting the local ordering of dimensions; if accuracy collapses to random-pooling levels, the local-correlation assumption is the load-bearing premise.

Watch

Extended reading notes

Core claim

The central discovery is that word embeddings behave enough like signals with locally correlated dimensions that a fixed wavelet low-pass/high-pass split can separate salient semantic content from detail, and the resulting coefficients can be summarized across a sentence with DCT without needing learned parameters. The authors show that Level-1 approximation (cA) and detail (cD) coefficients, at half the dimension, match or exceed original embeddings on word-similarity tasks, and that combining them with DCT produces sentence vectors of dimension d (the word-embedding size) that outperform DCT-only vectors of dimension Kd and even match or beat averaging on most tasks. They also report qualitative evidence that low-pass coefficients capture category-level commonalities such as animals or emotions, while high-pass coefficients preserve contrasts such as gender or age.

Load-bearing premise

The method assumes that the dimensions of a word embedding are locally correlated with their neighbors in the vector, so that fixed wavelet low-pass filters that average nearby coordinates preserve the semantics needed for downstream tasks; the paper states it does not prove this correlation.

Editorial extensions

If this is right

  • Sentence embedding size stays at the word-embedding dimension $d$ instead of growing as $Kd$, where $K$ is the number of DCT coefficients, so downstream memory and computation drop sharply for the same information.
  • Because the transform is parameter-free, it applies directly to any pretrained word embeddings without training, making it a drop-in compression for existing pipelines.
  • DWT alone compresses word embeddings to 50–75% of their original size while preserving word-similarity and categorization performance, which is useful when storage or retrieval speed is the bottleneck.
  • On several tasks (CR, SST2, MRPC, STS12), the DWT-DCT model outperforms both averaging and DCT-only baselines, indicating that the two transforms capture complementary information rather than redundant signal.

Reading between the lines

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

  • The paper does not compare against learned pooling operators of the same size; a natural next test would be to replace the fixed wavelet filters with a small learned 1-D convolution, which would show whether the fixed basis is near-optimal or leaves performance on the table.
  • Because the method fixes the sentence vector to the word dimension, it could be used as a cheap input projection for long-document or sequence models, potentially reducing the cost of self-attention over many tokens.
  • The qualitative split between approximation coefficients (which capture category-level similarity) and detail coefficients (which preserve contrasts) suggests that a task-specific weighting of the two channels could improve results further, an option the paper does not explore.
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

4 major / 4 minor

Summary. The paper proposes using the Discrete Wavelet Transform (DWT) to compress word embeddings along their dimension axis, and combining DWT with the Discrete Cosine Transform (DCT) to produce fixed-size sentence embeddings. The method is deterministic and non-parameterized: it applies a mother wavelet row-wise to the sentence embedding matrix, then applies DCT column-wise over the resulting coefficient matrices, keeping one or two DCT coefficients. The authors evaluate the word-level embeddings on similarity and categorization tasks, and the sentence-level embeddings on SentEval classification and semantic textual similarity tasks, comparing against averaging, DCT, EigenSent, p-means, and VLAME baselines. They report that DWT-DCT matches or improves on DCT-only sentence embeddings at substantially lower dimensionality.

Significance. The core idea is simple, novel in combination, and, if validated, practically useful: a fixed-size sentence embedding that is cheap, deterministic, and does not require training. The paper's breadth of evaluation (word similarity, categorization, and nine downstream tasks) and the qualitative analyses are assets. However, the manuscript currently does not supply a reproducible configuration (the mother wavelet and scale are selected per task without a validation protocol), reports no variance or significance measures, and relies on external baseline numbers for the central comparison. These issues must be resolved before the central claim of 'comparable and even superior' performance can be accepted.

major comments (4)
  1. [§6.1, Tables 6 and 7] The central comparison in Table 7 is not based on a fixed configuration. Section 6.1 states that 'the specific details of MW and its application in each task' are omitted, and Table 6's caption indicates that the mother wavelet used per condition per task is shown in italics. Because the method is presented as non-parameterized (Section 3), the per-task choice of wavelet family, scale, number of DWT levels L, and the K=2 skip trick are effectively free choices made after seeing the results. Without a pre-registered or nested validation protocol that fixes these choices on development data only, the reported gains (e.g., CR 81.14 vs 79.81 in Table 7) may reflect test-set selection. Please report the exact wavelet configuration per task and a validation-based selection procedure.
  2. [§4.3 and §5.1, Tables 6 and 7] No error bars, confidence intervals, or significance tests are reported for any of the tables. The differences that support the headline claims are small: in Table 7, DWT1-DCT[0] beats DCT c[0] by 1.33 points on CR and 1.15 on SST2, while being 0.07 points worse on MR. Since SentEval classifiers are stochastic, these gaps are likely within run-to-run noise. In addition, the DCT baseline rows in Table 7 are taken from [2] ('as reported in [2]') rather than re-run in the same codebase, so differences in tokenization, classifier hyperparameters, or evaluation versions are uncontrolled. Please re-run all baselines in the same pipeline and report variance or bootstrap confidence intervals.
  3. [Table 6] Table 6 contains corrupted entries that prevent interpretation. The row A VG(cD) contains the value '76.61coif15' in the MR column, and the A VG(Random Pooling) row contains '8' for TREC and '33' for STS12, which are implausible accuracy and Pearson correlation values. These entries must be corrected or explained, and the table caption should be checked for the repeated model name 'A VG(cD+cAD+cAAD)'.
  4. [§6.2] The paper admits that 'we did not explicitly prove that the dimensions in the same embedding are correlated.' This correlation is the premise for applying a wavelet filter along the index order of embedding dimensions. The argument that DWT achieves comparable results therefore presupposes the conclusion. Please add a direct test, for example comparing DWT compression to the same transform applied to randomly permuted embedding dimensions and to a random projection of the same output dimension. Without such a control, the proposed mechanism for why DWT helps is not established.
minor comments (4)
  1. [§3] The description of the DWT coefficient matrices is confusing: for L=2, the text lists cDD, cDA, cAD, and cAA, but Section 4.2 uses 'cD+cAD' and 'cA+cDA' without defining how these matrices are concatenated. Please clarify the naming and the concatenation order.
  2. [§6] The conclusion refers to 'DCT[0:2]' but Table 7 only shows DCT c[0:1]. Please align the notation between the text and the table.
  3. [§4.3] The text states that averaging (AVG) with FastText embeddings is the baseline, but Table 6 also lists 'A VG(Random Pooling)' and 'A VG(PCA+PPA)'. Please clarify whether all baselines use FastText and how the random pooling condition was constructed.
  4. [Tables 1, 6, 7] The captions say that the best results are shown in red, but color is not visible in a printed or grayscale copy. Please use bold or another typographic distinction that survives monochrome printing.

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation-level circularity; the only self-citation concern is the DCT baseline from [2], which is not load-bearing.

  1. other [Section 3 (K-coefficient choice) and Table 7 caption (DCT baseline source)]
    "We finally encode a sentence vector of size d by concatenating either 1 or 2 DCT coefficients, which proved to be sufficient empirically and also consistent with observations in [2]. ... Results as opposed to DCT embedding, DCT c*, as reported in [2]."

    The DCT baseline numbers and the K=1/2 coefficient convention are taken from [2], which shares an author (Mona Diab) with this paper. This is a self-citation signal, but it is not load-bearing: the DWT-DCT sentence vector is a fixed deterministic composition of DWT and DCT applied to the input word vectors, with no learned parameters and no target-label information entering the transform. The cited DCT values are an external benchmark, and the coefficient choice is a reused hyperparameter convention rather than a result forced by the transform definition. The reported DWT-DCT accuracies therefore do not reduce to this citation by construction; the concern is minor and primarily about benchmark provenance and reproducibility.

full rationale

The paper's central claim is empirical: applying a fixed Discrete Wavelet Transform row-wise to word vectors and then a Discrete Cosine Transform column-wise yields sentence embeddings that are competitive with or better than averaging and DCT baselines on SentEval tasks. The transform itself is a deterministic, non-parameterized function of the input embeddings, so the reported results are not equivalent to an input by construction and no fitted parameter is renamed as a prediction. The only same-group citation is [2], which supplies the DCT baseline values in Table 7 and the K=1/2 coefficient convention; this is benchmark reuse and a hyperparameter convention, not a load-bearing derivation. Section 6.2 explicitly concedes that dimension correlation is not proven, and Section 6.1 admits that per-task mother-wavelet details are omitted; these are reproducibility and validation limitations, not circular reasoning. Thus the derivation chain is self-contained against external benchmarks, with a minor non-load-bearing self-citation.

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

The central method is deterministic, but it rests on a domain assumption about the spatial structure of word embeddings and on several hyperparameters (wavelet family, scale, number of levels, DCT coefficient count) that are selected per task. These are listed as free parameters. No new entities are introduced.

free parameters (4)
  • Mother wavelet family and scale = Varies per task, e.g., 'coif15' in Table 6
    Chosen per task to maximize downstream performance, not predetermined.
  • Number of DWT levels L = 1, 2, or 3 depending on task
    Selected per task; Levels 1-3 appear in Tables 6 and 7.
  • Number of DCT coefficients K = 1 or 2
    The paper uses 1 or 2 coefficients, having found them sufficient empirically (Section 3).
  • DWT coefficient skip factor for K=2 = Skip every other coefficient
    An ad hoc trick (Section 3) to keep the final sentence vector at dimension d; its information preservation is only empirically asserted.
assumptions (3)
  • domain assumption Word embedding dimensions are sufficiently correlated along their index ordering for a fixed wavelet low-pass filter to preserve semantic information.
    The paper's compression strategy assumes neighboring dimensions carry redundant or locally structured information; the authors admit in Section 6.2 that this correlation is not proven.
  • domain assumption The selected mother wavelet families (Coiflets, Daubechies, Symlets) are appropriate for the structure of embedding data.
    Wavelet choice is made empirically per task without a theoretical justification, as described in Section 6.1.
  • domain assumption DCT coefficients over word positions capture sentence-level semantic patterns useful for downstream tasks.
    This assumption is borrowed from the DCT sentence embedding line of work [2] and is not re-derived here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Combining Discrete Wavelet and Cosine Transforms for Efficient Sentence Embedding." pith.science (2026). https://pith.science/paper/ZEHBHLMY

@misc{pith2026250800420,
  author       = {Pith},
  title        = {Pith review of: Combining Discrete Wavelet and Cosine Transforms for Efficient Sentence Embedding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZEHBHLMY}},
  note         = {Machine review of arXiv:2508.00420}
}
read the original abstract

Wavelets have emerged as a cutting edge technology in a number of fields. Concrete results of their application in Image and Signal processing suggest that wavelets can be effectively applied to Natural Language Processing (NLP) tasks that capture a variety of linguistic properties. In this paper, we leverage the power of applying Discrete Wavelet Transforms (DWT) to word and sentence embeddings. We first evaluate, intrinsically and extrinsically, how wavelets can effectively be used to consolidate important information in a word vector while reducing its dimensionality. We further combine DWT with Discrete Cosine Transform (DCT) to propose a non-parameterized model that compresses a sentence with a dense amount of information in a fixed size vector based on locally varying word features. We show the efficacy of the proposed paradigm on downstream applications models yielding comparable and even superior (in some tasks) results to original embeddings.

Figures

Figures reproduced from arXiv: 2508.00420 by the authors.

Figure 1
Figure 1. An example of Level-1 Wavelet transform using DWT to the cameraman image (1) and the corresponding transformed image (2) The transform, (2), produces 4 sub-bands. The top left sub-band is a low resolution version of the original sub-sampled in both horizontal and vertical directions, while the horizontal, h, vertical, v, and diagonal, d, details represent the down-sampled residual versions of the original image. bot… view at source ↗
Figure 2
Figure 2. Plotting word embedding of the word ‘work’ and its coefficients across 2 levels of WT. cD and cA represent Detail and Approximation coefficients derived from Level-1 of the transform. cDA and cDD indicate Level-2 Approximation and Detail coefficients derived from applying DWT on the Level-1 Detail coefficients. cAA and cAD indicate Level-2 Approximation and Detail coefficients derived from applying DWT on the Level-… view at source ↗
Figure 3
Figure 3. Proposed Wavelet-DCT sentence embedding model applied to a sentence of 4 words rep￾resented as a matrix. Level-1 DWT and DCT-1 are used as demonstrated on the left side, while the right side demonstrates using Level-1 DWT with DCT-2 by skipping every-other coefficient in DWT coefficients matrix to generate sentence embedding representation. 4 For a more detailed explanation of WT theory, refer to [13] [27] [9]. 6 [… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

44 extracted references · 40 canonical work pages

  1. [2]

    Efficient sentence embedding using dis- crete cosine transform

    Nada Almarwani, Hanan Aldarmaki, and Mona Diab. Efficient sentence embedding using dis- crete cosine transform. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Process- ing (EMNLP-IJCNLP), pages 3672–3678, Hong Kong, China, November 2019. Association f...

    work page 2019

  2. [1]

    SemEval-2012 task 6: A pilot on semantic textual similarity

    Eneko Agirre, Daniel Cer, Mona Diab, and Aitor Gonzalez-Agirre. SemEval-2012 task 6: A pilot on semantic textual similarity. In *SEM 2012: The First Joint Conference on Lexical and Computational Semantics – Volume 1: Proceedings of the main conference and the shared task, and Volume 2: Proceedings of the Sixth International Workshop on Semantic Evaluation...

    work page 2012

  3. [3]

    Attributes in lexical acquisition

    Abdulrahman Almuhareb. Attributes in lexical acquisition. 2006

    work page 2006

  4. [4]

    A simple but tough-to-beat baseline for sen- tence embeddings

    Sanjeev Arora, Yingyu Liang, and Tengyu Ma. A simple but tough-to-beat baseline for sen- tence embeddings. January 2019. 5th International Conference on Learning Representations, ICLR 2017 ; Conference date: 24-04-2017 Through 26-04-2017

    work page 2019

  5. [5]

    Don’t count, predict! a systematic comparison of context-counting vs

    Marco Baroni, Georgiana Dinu, and Germ´ an Kruszewski. Don’t count, predict! a systematic comparison of context-counting vs. context-predicting semantic vectors. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 238–247, Baltimore, Maryland, June 2014. Association for Computational Li...

    work page 2014

  6. [6]

    How we blessed distributional semantic evaluation

    Marco Baroni and Alessandro Lenci. How we blessed distributional semantic evaluation. pages 1–10, 07 2011

    work page 2011

  7. [7]

    The Wavelet Transform for Image Processing Appli- cations

    Toufik Bouden and Mokhtar Nibouche. The Wavelet Transform for Image Processing Appli- cations. 04 2012

    work page 2012

  8. [8]

    Multimodal distributional semantics

    Elia Bruni, Nam Khanh Tran, and Marco Baroni. Multimodal distributional semantics. J. Artif. Int. Res., 49(1):1–47, January 2014

    work page 2014

Show all 44 references
  1. [9]

    Brunton and J

    Steven L. Brunton and J. Nathan Kutz. Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control. Cambridge University Press, 2019

    2019

  2. [10]

    Signal processing and compression with wavelet packets

    Ronald Coifman, Yves Meyer, Steven Quake, and Mladen Wickerhauser. Signal processing and compression with wavelet packets. Tech rep, 01 1991

    1991

  3. [11]

    Senteval: An evaluation toolkit for universal sentence representations

    Alexis Conneau and Douwe Kiela. Senteval: An evaluation toolkit for universal sentence representations. CoRR, abs/1803.05449, 2018

    2018 arXiv

  4. [12]

    A wavelet-packets based algorithm for eeg signal compression

    Juli´ an L C´ ardenas-Barrera, Juan V Lorenzo-Ginori, and Ernesto Rodr ´ ıguez-Valdivia. A wavelet-packets based algorithm for eeg signal compression. Medical Informatics and the Internet in Medicine, 29(1):15–27, 2004

    2004

  5. [13]

    Ten Lectures on Wavelets

    Ingrid Daubechies. Ten Lectures on Wavelets. Society for Industrial and Applied Mathematics, USA, 1992

    1992

  6. [14]

    C ¸ . P. Dautov and M. S. ¨Ozerdem. Wavelet transform and signal denoising using wavelet method. In 2018 26th Signal Processing and Communications Applications Conference (SIU), pages 1–4, 2018

    2018

  7. [15]

    Unsupervised construction of large paraphrase corpora: Exploiting massively parallel news sources

    Bill Dolan, Chris Quirk, and Chris Brockett. Unsupervised construction of large paraphrase corpora: Exploiting massively parallel news sources. In COLING 2004: Proceedings of the 20th International Conference on Computational Linguistics, pages 350–356, Geneva, Switzerland, au...

    2004

  8. [16]

    Placing search in context: The concept revisited

    Lev Finkelstein, Evgeniy Gabrilovich, Yossi Matias, Ehud Rivlin, Zach Solan, Gadi Wolfman, and Eytan Ruppin. Placing search in context: The concept revisited. volume 20, pages 406– 414, 01 2001

    2001

  9. [17]

    A. Graps. An introduction to wavelets. 1995

    1995

  10. [18]

    Grgic, M

    S. Grgic, M. Grgic, and B. Zovko-Cihlar. Performance analysis of image compression using wavelets. IEEE Transactions on Industrial Electronics, 48(3):682–695, 2001

    2001

  11. [19]

    Recent advances in convolutional neural networks

    Jiuxiang Gu, Zhenhua Wang, Jason Kuen, Lianyang Ma, Amir Shahroudy, Bing Shuai, Ting Liu, Xingxing Wang, and Gang Wang. Recent advances in convolutional neural networks. CoRR, abs/1512.07108, 2015

    2015 arXiv

  12. [20]

    Simlex-999: Evaluating semantic models with (genuine) similarity estimation

    Felix Hill, Roi Reichart, and Anna Korhonen. Simlex-999: Evaluating semantic models with (genuine) similarity estimation. CoRR, abs/1408.3456, 2014

    2014 arXiv

  13. [21]

    Mining and summarizing customer reviews

    Minqing Hu and Bing Liu. Mining and summarizing customer reviews. In KDD ’04, 2004

    2004

  14. [22]

    Radu Tudor Ionescu and Andrei M. Butnaru. Vector of locally-aggregated word embeddings (vlawe): A novel document-level representation, 2019

    2019

  15. [23]

    EigenSent: Spectral sentence embeddings using higher-order dynamic mode decomposition

    Subhradeep Kayal and George Tsatsaronis. EigenSent: Spectral sentence embeddings using higher-order dynamic mode decomposition. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4536–4546, Florence, Italy, July 2019. Association ...

    2019

  16. [24]

    A. S. Lewis and G. Knowles. Image compression using the 2-d wavelet transform. IEEE Transactions on Image Processing, 1(2):244–250, 1992

    1992

  17. [25]

    A survey on wavelet applications in data mining

    Tao Li, Qi Li, Shenghuo Zhu, and Mitsunori Ogihara. A survey on wavelet applications in data mining. SIGKDD Explorations, 4:49–68, 12 2002

    2002

  18. [26]

    Word Embedding for Understanding Natural Language: A Survey, volume 26

    Yang Li and Tao Yang. Word Embedding for Understanding Natural Language: A Survey, volume 26. 05 2017

    2017

  19. [27]

    Madhavan

    G. Madhavan. The illustrated wavelet transform handbook - introductory theory and ap- plications in science, engineering, medicine and finance [book review]. IEEE Engineering in Medicine and Biology Magazine, 22(1):92–93, 2003. 17

    2003

  20. [28]

    Advances in pre-training distributed word representations

    Tomas Mikolov, Edouard Grave, Piotr Bojanowski, Christian Puhrsch, and Armand Joulin. Advances in pre-training distributed word representations. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan, May 2018....

    2018

  21. [29]

    Linguistic regularities in continuous space word representations

    Tomas Mikolov, Wen-tau Yih, and Geoffrey Zweig. Linguistic regularities in continuous space word representations. In Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 746–751, A...

    2013

  22. [30]

    All-but-the-top: Simple and effective postpro- cessing for word representations

    Jiaqi Mu, Suma Bhat, and Pramod Viswanath. All-but-the-top: Simple and effective postpro- cessing for word representations. CoRR, abs/1702.01417, 2017

    2017 arXiv

  23. [31]

    Selecting corpus-semantic models for neurolinguistic decoding

    Brian Murphy, Partha Talukdar, and Tom Mitchell. Selecting corpus-semantic models for neurolinguistic decoding. In Proceedings of the First Joint Conference on Lexical and Compu- tational Semantics - Volume 1: Proceedings of the Main Conference and the Shared Task, and Volume ...

    2012

  24. [33]

    A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts

    Bo Pang and Lillian Lee. A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. In Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL-04), pages 271–278, Barcelona, Spain, July 2004

    2004

  25. [34]

    GloVe: Global vectors for word representation

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

    2014

  26. [35]

    Effective dimensionality reduction for word embeddings

    Vikas Raunak, Vivek Gupta, and Florian Metze. Effective dimensionality reduction for word embeddings. In Proceedings of the 4th Workshop on Representation Learning for NLP (RepL4NLP-2019), pages 235–243, Florence, Italy, August 2019. Association for Computa- tional Linguistics

    2019

  27. [36]

    Rioul and M

    O. Rioul and M. Vetterli. Wavelets and signal processing. IEEE Signal Processing Magazine, 8(4):14–38, 1991

    1991

  28. [37]

    Concatenated p-mean word embeddings as universal cross-lingual sentence representations

    Andreas R¨ uckl´ e, Steffen Eger, Maxime Peyrard, and Iryna Gurevych. Concatenated p-mean word embeddings as universal cross-lingual sentence representations. CoRR, abs/1803.01400, 2018

    2018 arXiv

  29. [38]

    Chapter 10 - compression

    Mallat St´ ephane. Chapter 10 - compression. In Mallat St´ ephane, editor, A Wavelet Tour of Signal Processing (Third Edition), pages 481 – 533. Academic Press, Boston, third edition edition, 2009

    2009

  30. [39]

    Speech recognition by wavelet analysis

    Nitin Trivedi, Vikesh Kumar, Saurabh Singh, Sachin Ahuja, and Raman Chadha. Speech recognition by wavelet analysis. International Journal of Computer Applications, 15:27–32, 2011

    2011

  31. [40]

    Voorhees and Dawn M

    Ellen M. Voorhees and Dawn M. Tice. Building a question answering test collection. In Proceedings of the 23rd Annual International ACM SIGIR Conference on Research and De- velopment in Information Retrieval, SIGIR ’00, page 200–207, New York, NY, USA, 2000. Association for Com...

    2000

  32. [41]

    Bin Wang, Angela Wang, Fenxiao Chen, Yuncheng Wang, and C.-C. Jay Kuo. Evaluating word embedding models: methods and experimental results. APSIPA Transactions on Signal and Information Processing, 8:e19, 2019

    2019

  33. [42]

    On the dimensionality of sentence embed- dings

    Hongwei Wang, Hongming Zhang, and Dong Yu. On the dimensionality of sentence embed- dings. In Houda Bouamor, Juan Pino, and Kalika Bali, editors, Findings of the Association 18 for Computational Linguistics: EMNLP 2023, pages 10344–10354, Singapore, December 2023. Association ...

    2023

  34. [43]

    Z. Xizhi. The application of wavelet transform in digital image processing. In 2008 Interna- tional Conference on MultiMedia and Information Technology, pages 326–329, 2008

    2008

  35. [44]

    An edge detection approach based on directional wavelet transform

    Zhen Zhang, Siliang Ma, Hui Liu, and Yuexin Gong. An edge detection approach based on directional wavelet transform. Computers and Mathematics with Applications, 57(8):1265 – 1271, 2009

    2009

  36. [45]

    Sentence analogies: Exploring linguistic relationships and regularities in sentence embeddings, 2020

    Xunjie Zhu and Gerard de Melo. Sentence analogies: Exploring linguistic relationships and regularities in sentence embeddings, 2020. 19

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.