Pith. sign in

REVIEW 3 major objections 1 minor 98 references

FlexMUSE: Multimodal Unification and Semantics Enhancement Framework with Flexible interaction for Creative Writing

T0 review · 3 major / 1 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The FlexMUSE abstract claims a creative-writing framework; the supplied body is a different paper.

desk verdict The submitted PDF is not the FlexMUSE paper; it is the UMATO dimensionality reduction manuscript, so the abstract's claims about FlexMUSE have zero support in the body. read the letter →

arxiv 2508.16230 v1 pith:AGBCDALN submitted 2025-08-22 cs.CV cs.AI

Jiahao Chen , Zhiyong Ma , Wenbiao Du , Qingyuan Chuai This is my paper · ORCID
classification cs.CVcs.AI
keywords FlexMUSEmulti-modalcreativewritingmsaGatemscDPOArtUMATOdimensionalityreductionmanuscriptmismatch
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

This submission's abstract claims FlexMUSE, a framework for multi-modal creative writing—producing illustrated articles where text and images are semantically aligned. It introduces msaGate, attention-based cross-modality fusion, and mscDPO training, along with the ArtMUSE dataset of about 3,000 calibrated text-image pairs, and reports gains in consistency, creativity, and coherence. The full text supplied with the submission, however, is not this paper: it is UMATO, a two-phase dimensionality-reduction method for visual analytics. None of the FlexMUSE components, dataset, or creative-writing experiments appears in the body. A sympathetic reader can therefore state the abstract's claim, but the supporting manuscript does not contain it.

What carries the argument

For the abstract's claim, the load-bearing machinery is msaGate (modality semantic alignment gating, which restricts textual input to align it with visual semantics), an attention-based cross-modality fusion that augments input features, mscDPO (a variant of direct preference optimization whose rejected samples are extended), and the ArtMUSE dataset of roughly 3,000 calibrated text-image pairs. For the body as actually supplied, the machinery is UMATO's two-phase optimization: representative hub points are selected by kNN frequency and laid out first without negative-sampling approximation to fix global structure, then expanded nearest neighbors are embedded with UMAP's local objective while

What would settle it

Open the supplied full text and search for msaGate, mscDPO, or ArtMUSE: none appears. A reader could also check whether any experiment in the body concerns creative writing rather than dimensionality-reduction scatterplots; reproducing the abstract's claimed results is impossible because the body contains only UMATO.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that multi-modal creative writing can be done economically with flexible interaction: a text-to-image module lets visual input be optional, a gating mechanism (msaGate) restricts textual input to align modalities, an attention-based fusion augments input features, and a modified direct preference optimization (mscDPO) extends rejected samples to push creativity. The abstract asserts that these components together improve consistency, creativity, and coherence. The body actually supplied is a separate manuscript about UMATO, a dimensionality-reduction technique that preserves both local and global structure by optimizing a skeleton of hub points first and th

Load-bearing premise

Read as a whole, the load-bearing premise—that the supplied full text is the FlexMUSE paper—fails on inspection; read as an abstract alone, the load-bearing premise is that about 3,000 calibrated text-image pairs suffice to demonstrate the claimed alignment and creativity.

Editorial extensions

If this is right

  • If the abstract's claim holds, illustrated-article generation could accept optional image input and still keep text and image semantics aligned, without costly per-task training.
  • If msaGate works as intended, the same model should produce writing whose entities and events match a given image even when the image is only loosely related to the text prompt.
  • If mscDPO genuinely enhances creativity, generated articles should show greater lexical and structural variation than a baseline trained with standard preference optimization, while staying coherent.
  • If ArtMUSE is a usable benchmark, future multi-modal creative-writing systems could be compared on the same roughly 3,000 calibrated pairs.
  • These corollaries follow from the abstract alone; the supplied body supplies no evidence for them.

Reading between the lines

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

  • A direct ablation test would settle the abstract's design claim: remove msaGate and measure cross-modal semantic alignment, and remove mscDPO and measure creative divergence, on the same output pairs.
  • The UMATO body, taken on its own, suggests a transferable design principle—optimize global structure first on a small skeleton, then fill in local detail—that could be tested in other optimization-based embedding tasks beyond two-dimensional projection.
  • The ArtMUSE dataset's 'calibrated' construction is unspecified; a useful check would be to compare a model trained on ArtMUSE with a model trained on an equal-sized sample of existing image-text data to see whether the calibration, not just the size, drives the reported alignment.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 1 minor

Summary. The submission, arXiv:2508.16230, presents an abstract for a framework called FlexMUSE for multi-modal creative writing (MMCW). The abstract claims: (i) FlexMUSE introduces a T2I module for optional visual input; (ii) it uses a modality semantic alignment gating (msaGate) to restrict textual input; (iii) it proposes attention-based cross-modality fusion; (iv) it introduces modality semantic creative direct preference optimization (mscDPO); (v) it introduces an ArtMUSE dataset of about 3,000 calibrated text-image pairs; and (vi) it achieves promising results in consistency, creativity, and coherence. The full text supplied with the submission, however, is a completely different paper: 'UMATO: Bridging Local and Global Structures for Reliable Visual Analytics with Dimensionality Reduction' by Jeon et al., carrying the arXiv header arXiv:2508.16227v1 [cs.LG]. The body contains no description of FlexMUSE, no msaGate definition, no mscDPO objective, no ArtMUSE dataset description, no experiments, and no results related to creative writing or image-text alignment. Every substantive claim in the FlexMUSE abstract is therefore unsupported by the document submitted for review.

Significance. If FlexMUSE were correctly presented, the paper could be relevant to multimodal generation, proposing a new task formulation (MMCW), a gating mechanism for semantic alignment, a preference-optimization variant, and a new dataset. However, none of these contributions appear in the submitted manuscript. The actual full text is the UMATO paper, which addresses dimensionality reduction for visual analytics. That paper may be a legitimate contribution to its own field, but it shares no methods, experiments, or results with the FlexMUSE abstract. Because the submitted document does not contain the claimed framework, dataset, or evaluation, the central claim of the abstract—that FlexMUSE achieves promising results—has no supporting content in the manuscript. The paper cannot be meaningfully evaluated for correctness, novelty, or reproducibility as submitted.

major comments (3)
  1. [Full text (all sections)] The submitted full text is not the FlexMUSE paper. The abstract describes a multimodal creative-writing framework with msaGate, attention-based cross-modality fusion, mscDPO, and the ArtMUSE dataset, but the body is the UMATO paper on dimensionality reduction (arXiv:2508.16227v1). There is no architecture, no training objective, no dataset description, no experimental setup, and no results for FlexMUSE anywhere in the manuscript. This is not a missing derivation or a weak evaluation; it is a complete absence of the claimed content.
  2. [Abstract, final sentence] The only load-bearing evidence statement in the abstract is 'FlexMUSE achieves promising results, demonstrating its consistency, creativity and coherence.' The supplied manuscript contains no quantitative results, no baselines, no ablation studies, no generated samples, and no evaluation protocol for creativity, consistency, or coherence. The claim is therefore unsupported by the submitted document.
  3. [Abstract, ArtMUSE dataset] The abstract claims a dataset of 'around 3k calibrated text-image pairs.' The manuscript does not describe the dataset, the meaning of 'calibrated,' the collection procedure, or any use of the dataset in experiments. Without this information, the dataset contribution cannot be assessed.
minor comments (1)
  1. [General] If the intended paper is the UMATO manuscript, the submitted abstract and title are incorrect and should be replaced. If the intended paper is FlexMUSE, the correct full text must be uploaded. Neither the title, abstract, nor body can be reconciled as a single coherent submission.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation can be found: the manuscript body is a different paper (UMATO), so FlexMUSE's abstract claims have no in-body derivation chain to audit.

full rationale

The submitted document's abstract describes FlexMUSE, msaGate, mscDPO, ArtMUSE, and promising multimodal creative-writing results, but the full text is entirely the UMATO paper by Jeon et al. (IEEE TVCG, arXiv:2508.16227), with no FlexMUSE architecture, dataset description, training objective, or experiments. Because none of the abstract's claimed components appear in the body, there is no derivation chain whose steps could be shown to reduce to their own inputs. The only candidate circularity—the abstract's claims resting on missing content—is not a reduction by construction or by self-citation; it is an absence of evidence, which is a soundness/review-integrity problem rather than a circularity finding. The UMATO body itself is a conventional empirical paper: it builds on UMAP's published loss functions (Eq. 5 and 8), proposes a two-phase optimization, and validates against 20 external real-world datasets (Table 1) and external baselines using established metrics (T&C, MRREs, S&C, KL divergence, DTM, Stress). Its self-citations (e.g., the VIS 2022 short paper [22] and prior metric papers [15,16,73]) are used for context and measurement, not to force the empirical results. No fitted parameter is renamed as a prediction, and no uniqueness theorem or ansatz is imported from the authors' prior work in a load-bearing way. Therefore the correct circularity score is 0: there is no circular step to exhibit, and the dominant issue is the abstract/body mismatch, not circular reasoning.

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

No free parameters can be audited because the manuscript body does not contain the FlexMUSE method. The axioms and entities above are extracted from the abstract alone. All three invented entities (msaGate, mscDPO, ArtMUSE) carry no independent evidence within this submission and are not described in the body text.

assumptions (2)
  • domain assumption Multi-modal creative writing is a well-defined task in which generated text and images should be semantically aligned even though their contexts are "not strictly related" (abstract).
    The abstract's problem framing presupposes a measurable notion of semantic alignment between loosely related modalities, but no metric or formal definition is given anywhere in the document.
  • domain assumption Direct preference optimization can be extended to creative writing by expanding rejected samples (mscDPO).
    The abstract posits that DPO-style preference learning transfers to creative writing without showing the preference data construction, reward model, or training procedure.
invented entities (3)
  • msaGate (modality semantic alignment gating)
    purpose: Restrict the textual input so that generated text and images stay semantically aligned during multi-modal creative writing.
    Described only in the abstract; no mechanism, equations, or ablations appear in the manuscript body.
  • mscDPO (modality semantic creative direct preference optimization)
    purpose: Train the framework for writing creativity by extending the rejected samples in direct preference optimization.
    No training details or quantitative effect are provided; the full text contains no reference to this objective.
  • ArtMUSE dataset
    purpose: Benchmark for multi-modal creative writing with about 3,000 calibrated text-image pairs.
    Announced in the abstract only; no access link, statistics, collection protocol, or samples are provided, and the body text does not describe it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FlexMUSE: Multimodal Unification and Semantics Enhancement Framework with Flexible interaction for Creative Writing." pith.science (2026). https://pith.science/paper/AGBCDALN

@misc{pith2026250816230,
  author       = {Pith},
  title        = {Pith review of: FlexMUSE: Multimodal Unification and Semantics Enhancement Framework with Flexible interaction for Creative Writing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AGBCDALN}},
  note         = {Machine review of arXiv:2508.16230}
}
read the original abstract

Multi-modal creative writing (MMCW) aims to produce illustrated articles. Unlike common multi-modal generative (MMG) tasks such as storytelling or caption generation, MMCW is an entirely new and more abstract challenge where textual and visual contexts are not strictly related to each other. Existing methods for related tasks can be forcibly migrated to this track, but they require specific modality inputs or costly training, and often suffer from semantic inconsistencies between modalities. Therefore, the main challenge lies in economically performing MMCW with flexible interactive patterns, where the semantics between the modalities of the output are more aligned. In this work, we propose FlexMUSE with a T2I module to enable optional visual input. FlexMUSE promotes creativity and emphasizes the unification between modalities by proposing the modality semantic alignment gating (msaGate) to restrict the textual input. Besides, an attention-based cross-modality fusion is proposed to augment the input features for semantic enhancement. The modality semantic creative direct preference optimization (mscDPO) within FlexMUSE is designed by extending the rejected samples to facilitate the writing creativity. Moreover, to advance the MMCW, we expose a dataset called ArtMUSE which contains with around 3k calibrated text-image pairs. FlexMUSE achieves promising results, demonstrating its consistency, creativity and coherence.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

98 extracted references · 70 canonical work pages

  1. [1]

    Panene: A progressive algorithm for indexing and querying approximate k-nearest neighbors,

    J. Jo, J. Seo, and J.-D. Fekete, “Panene: A progressive algorithm for indexing and querying approximate k-nearest neighbors,” IEEE Trans- actions on Visualization and Computer Graphics , vol. 26, no. 2, pp. 1347–1360, 2018

    2018

  2. [2]

    Supporting analysis of dimen- sionality reduction results with contrastive learning,

    T. Fujiwara, O.-H. Kwon, and K.-L. Ma, “Supporting analysis of dimen- sionality reduction results with contrastive learning,” IEEE Transactions on Visualization and Computer Graphics, vol. 26, no. 1, pp. 45–55, 2019

    2019

  3. [3]

    t-visne: Interactive assessment and interpretation of t-sne projections,

    A. Chatzimparmpas, R. M. Martins, and A. Kerren, “t-visne: Interactive assessment and interpretation of t-sne projections,” IEEE Transactions on Visualization and Computer Graphics, vol. 26, no. 8, pp. 2696–2714, 2020

    2020

  4. [4]

    Dimensionality reduction for visualizing single-cell data using umap,

    E. Becht, L. McInnes, J. Healy, C.-A. Dutertre, I. W. Kwok, L. G. Ng, F. Ginhoux, and E. W. Newell, “Dimensionality reduction for visualizing single-cell data using umap,” Nature biotechnology, vol. 37, no. 1, pp. 38–44, 2019

    2019

  5. [5]

    Embedding comparator: Visualizing differences in global structure and local neighborhoods via small multiples,

    A. Boggust, B. Carter, and A. Satyanarayan, “Embedding comparator: Visualizing differences in global structure and local neighborhoods via small multiples,” in Proceedings of the 27th International Conference on Intelligent User Interfaces , ser. IUI ’22. New York, NY , USA: Association for Computing Machinery, 2022, p. 746–766

    2022

  6. [6]

    Multidimensional projection for visual analytics: Linking techniques with distortions, tasks, and layout enrich- ment,

    L. G. Nonato and M. Aupetit, “Multidimensional projection for visual analytics: Linking techniques with distortions, tasks, and layout enrich- ment,” IEEE Transactions on Visualization and Computer Graphics , vol. 25, no. 8, pp. 2650–2673, 2018

    2018

  7. [7]

    Revisit- ing dimensionality reduction techniques for visual cluster analysis: An empirical study,

    J. Xia, Y . Zhang, J. Song, Y . Chen, Y . Wang, and S. Liu, “Revisit- ing dimensionality reduction techniques for visual cluster analysis: An empirical study,” IEEE Transactions on Visualization and Computer Graphics, vol. 28, no. 1, pp. 529–539, 2021

    2021

  8. [8]

    Global versus local methods in nonlinear dimensionality reduction,

    V . D. Silva and J. B. Tenenbaum, “Global versus local methods in nonlinear dimensionality reduction,” in Advances in Neural Information Processing Systems, 2003, pp. 721–728

    2003

Show all 98 references
  1. [9]

    Umap: Uniform manifold approximation and projection for dimension reduction,

    L. McInnes, J. Healy, and J. Melville, “Umap: Uniform manifold approximation and projection for dimension reduction,” arXiv preprint arXiv:1802.03426, 2018

    2018 arXiv

  2. [10]

    Visualizing data using t-sne,

    L. v. d. Maaten and G. Hinton, “Visualizing data using t-sne,” Journal of machine learning research, vol. 9, no. Nov, pp. 2579–2605, 2008

    2008

  3. [11]

    Liii. on lines and planes of closest fit to systems of points in space,

    K. Pearson, “Liii. on lines and planes of closest fit to systems of points in space,” The London, Edinburgh, and Dublin philosophical magazine and journal of science, vol. 2, no. 11, pp. 559–572, 1901

    1901

  4. [12]

    A global geometric framework for nonlinear dimensionality reduction,

    J. B. Tenenbaum, V . De Silva, and J. C. Langford, “A global geometric framework for nonlinear dimensionality reduction,”Science, vol. 290, no. 5500, pp. 2319–2323, 2000

    2000

  5. [13]

    Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis,

    J. Kruskal, “Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis,” Psychometrika, vol. 29, pp. 1–27, 1964

    1964

  6. [14]

    Sparse multidimensional scaling using landmark points,

    V . De Silva and J. B. Tenenbaum, “Sparse multidimensional scaling using landmark points,” technical report, Stanford University, Tech. Rep., 2004

    2004

  7. [15]

    Classes are not clusters: Improving label-based evaluation of dimensionality reduction,

    H. Jeon, Y .-H. Kuo, M. Aupetit, K.-L. Ma, and J. Seo, “Classes are not clusters: Improving label-based evaluation of dimensionality reduction,” IEEE Transactions on Visualization and Computer Graphics , vol. 30, no. 1, pp. 781–791, 2024

    2024

  8. [16]

    Measuring and explain- ing the inter-cluster reliability of multidimensional projections,

    H. Jeon, H.-K. Ko, J. Jo, Y . Kim, and J. Seo, “Measuring and explain- ing the inter-cluster reliability of multidimensional projections,” IEEE Transactions on Visualization and Computer Graphics , vol. 28, no. 1, pp. 551–561, 2022

    2022

  9. [17]

    Feature learning for nonlinear dimensionality reduction toward maximal extraction of hidden patterns,

    T. Fujiwara, Y .-H. Kuo, A. Ynnerman, and K.-L. Ma, “Feature learning for nonlinear dimensionality reduction toward maximal extraction of hidden patterns,” in 2023 IEEE 16th Pacific Visualization Symposium (PacificVis), 2023, pp. 122–131

    2023

  10. [18]

    A critical analysis of the usage of dimensionality reduction in four domains,

    D. Cashman, M. Keller, H. Jeon, B. C. Kwon, and Q. Wang, “A critical analysis of the usage of dimensionality reduction in four domains,” IEEE Transactions on Visualization and Computer Graphics, pp. 1–20, 2025

    2025

  11. [19]

    The eyes have it: A task by data type taxonomy for information visualizations,

    B. Shneiderman, “The eyes have it: A task by data type taxonomy for information visualizations,” in Proceedings 1996 IEEE symposium on visual languages. IEEE, 1996, pp. 336–343

    1996

  12. [20]

    Understanding how dimension reduction tools work: An empirical approach to deciphering t-sne, umap, trimap, and pacmap for data visualization,

    Y . Wang, H. Huang, C. Rudin, and Y . Shaposhnik, “Understanding how dimension reduction tools work: An empirical approach to deciphering t-sne, umap, trimap, and pacmap for data visualization,” Journal of Machine Learning Research, vol. 22, no. 201, pp. 1–73, 2021

    2021

  13. [21]

    Trimap: Large-scale dimensionality reduction using triplets,

    E. Amid and M. K. Warmuth, “Trimap: Large-scale dimensionality reduction using triplets,” arXiv preprint arXiv:1910.00204, 2019

    1910 arXiv

  14. [22]

    Uniform manifold approximation with two-phase optimization,

    H. Jeon, H.-K. Ko, S. Lee, J. Jo, and J. Seo, “Uniform manifold approximation with two-phase optimization,” in 2022 IEEE Visualization and Visual Analytics (VIS), 2022, pp. 80–84

    2022

  15. [23]

    Appendix: Uniform manifold approximation with two-phase op- timization,

    ——, “Appendix: Uniform manifold approximation with two-phase op- timization,” 2022

    2022

  16. [24]

    Laplacian eigenmaps and spectral techniques for embedding and clustering,

    M. Belkin and P. Niyogi, “Laplacian eigenmaps and spectral techniques for embedding and clustering,” in Advances in Neural Information Processing Systems, 2002, pp. 585–591

    2002

  17. [25]

    Dis- tributed representations of words and phrases and their compositionality,

    T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean, “Dis- tributed representations of words and phrases and their compositionality,” in Advances in Neural Information Processing Systems, 2013, pp. 3111– 3119

    2013

  18. [26]

    Line: Large-scale information network embedding,

    J. Tang, M. Qu, M. Wang, M. Zhang, J. Yan, and Q. Mei, “Line: Large-scale information network embedding,” in Proceedings of the 24th International Conference on World Wide Web, 2015, pp. 1067–1077

    2015

  19. [27]

    Visualizing large-scale and high- dimensional data,

    J. Tang, J. Liu, M. Zhang, and Q. Mei, “Visualizing large-scale and high- dimensional data,” in Proceedings of the 25th International Conference on World Wide Web, 2016, pp. 287–297

    2016

  20. [28]

    Umap does not preserve global struc- ture any better than t-sne when using the same initialization,

    D. Kobak and G. C. Linderman, “Umap does not preserve global struc- ture any better than t-sne when using the same initialization,” BioRxiv, 2019

    2019

  21. [29]

    Checkviz: Sanity check and topological clues for linear and non-linear mappings,

    S. Lespinats and M. Aupetit, “Checkviz: Sanity check and topological clues for linear and non-linear mappings,” Computer Graphics Forum , vol. 30, no. 1, pp. 113–125, 2011

    2011

  22. [30]

    Distortion- aware brushing for interactive cluster analysis in multidimensional pro- jections,

    H. Jeon, M. Aupetit, S. Lee, H.-K. Ko, Y . Kim, and J. Seo, “Distortion- aware brushing for interactive cluster analysis in multidimensional pro- jections,” arXiv preprint arXiv:2201.06379, 2022

    2022

  23. [31]

    Neighborhood preservation in nonlinear projec- tion methods: An experimental study,

    J. Venna and S. Kaski, “Neighborhood preservation in nonlinear projec- tion methods: An experimental study,” in International Conference on Artificial Neural Networks. Springer, 2001, pp. 485–491

    2001

  24. [32]

    J. A. Lee and M. Verleysen, Nonlinear dimensionality reduction . Springer Science & Business Media, 2007

    2007

  25. [33]

    Stochastic neighbor embedding,

    G. E. Hinton and S. Roweis, “Stochastic neighbor embedding,” Advances in Neural Information Processing Systems, vol. 15, pp. 857–864, 2002

    2002

  26. [34]

    Large-scale evaluation of topic models and dimensionality reduction methods for 2d text spatialization,

    D. Atzberger, T. Cech, M. Trapp, R. Richter, W. Scheibel, J. D ¨ollner, and T. Schreck, “Large-scale evaluation of topic models and dimensionality reduction methods for 2d text spatialization,” IEEE Transactions on Visualization and Computer Graphics, vol. 30, no. 1, pp. 902–912, 2024

    2024

  27. [35]

    Towards a quantitative survey of dimension reduction techniques,

    M. Espadoto, R. M. Martins, A. Kerren, N. S. Hirata, and A. C. Telea, “Towards a quantitative survey of dimension reduction techniques,”IEEE Transactions on Visualization and Computer Graphics, 2019

    2019

  28. [36]

    Topological autoen- coders,

    M. Moor, M. Horn, B. Rieck, and K. Borgwardt, “Topological autoen- coders,” in Proceedings of the 37th International Conference on Machine Learning (ICML) , ser. Proceedings of Machine Learning Research. PMLR, 2020

    2020

  29. [37]

    Pattern trails: Visual analysis of pattern transitions in subspaces,

    D. J ¨ackle, M. Hund, M. Behrisch, D. A. Keim, and T. Schreck, “Pattern trails: Visual analysis of pattern transitions in subspaces,” in 2017 IEEE Conference on Visual Analytics Science and Technology (VAST) , 2017, pp. 1–12

    2017

  30. [38]

    Visual anal- ysis of dimensionality reduction quality for parameterized projections,

    R. M. Martins, D. B. Coimbra, R. Minghim, and A. Telea, “Visual anal- ysis of dimensionality reduction quality for parameterized projections,” Computers & Graphics, vol. 41, pp. 26–42, 2014

    2014

  31. [39]

    Visualizing distortions and recovering topology in contin- uous projection techniques,

    M. Aupetit, “Visualizing distortions and recovering topology in contin- uous projection techniques,” Neurocomputing, vol. 70, no. 7, pp. 1304– 1330, 2007, advances in Computational Intelligence and Learning

    2007

  32. [40]

    Uibert: Learning generic multimodal representations for ui understanding,

    C. Bai, X. Zang, Y . Xu, S. Sunkara, A. Rastogi, J. Chen, and B. A. y Arcas, “Uibert: Learning generic multimodal representations for ui understanding,” 2021

    2021

  33. [41]

    Representation selective self- distillation and wav2vec 2.0 feature exploration for spoof-aware speaker verification,

    J. W. Lee, E. Kim, J. Koo, and K. Lee, “Representation selective self- distillation and wav2vec 2.0 feature exploration for spoof-aware speaker verification,” arXiv preprint arXiv:2204.02639, 2022

    2022 arXiv

  34. [42]

    Can a computer tell differences between vibra- tions?: Physiology-based computational model for perceptual dissimilar- ity prediction,

    C. Lim and G. Park, “Can a computer tell differences between vibra- tions?: Physiology-based computational model for perceptual dissimilar- ity prediction,” in Proceedings of the 2023 CHI Conference on Human IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS 17 Factors i...

    2023

  35. [43]

    Vitality: Promoting serendipitous discovery of academic literature with transformers & visual analytics,

    A. Narechania, A. Karduni, R. Wesslen, and E. Wall, “Vitality: Promoting serendipitous discovery of academic literature with transformers & visual analytics,” IEEE Transactions on Visualization and Computer Graphics, vol. 28, no. 1, pp. 486–496, 2022

    2022

  36. [44]

    Visual analytics system of comprehensive data quality improvement for machine learning using data- and process-driven strategies,

    H. Hong, S. Yoo, Y . Jin, C. Yoon, S. Yim, S. Choi, and Y . Jang, “Visual analytics system of comprehensive data quality improvement for machine learning using data- and process-driven strategies,” in 2022 IEEE International Conference on Big Data (Big Data) , 2022, pp. 396– 401

    2022

  37. [45]

    Activis: Visual exploration of industry-scale deep neural network models,

    M. Kahng, P. Y . Andrews, A. Kalro, and D. H. Chau, “Activis: Visual exploration of industry-scale deep neural network models,” IEEE Trans- actions on Visualization and Computer Graphics , vol. 24, no. 1, pp. 88–97, 2018

    2018

  38. [46]

    Reducing the dimensionality of data with neural networks,

    G. E. Hinton and R. R. Salakhutdinov, “Reducing the dimensionality of data with neural networks,” Science, vol. 313, no. 5786, pp. 504–507, 2006

    2006

  39. [47]

    Local affine multidimensional projection,

    P. Joia, D. Coimbra, J. A. Cuminato, F. V . Paulovich, and L. G. Nonato, “Local affine multidimensional projection,” IEEE Transactions on Visualization and Computer Graphics, vol. 17, no. 12, pp. 2563–2571, 2011

    2011

  40. [48]

    Knowledge discovery on rfm model using bernoulli sequence,

    I.-C. Yeh, K.-J. Yang, and T.-M. Ting, “Knowledge discovery on rfm model using bernoulli sequence,” Expert Systems with Applications , vol. 36, no. 3, pp. 5866–5871, 2009

    2009

  41. [49]

    Deep learning classification in asteroseis- mology,

    M. Hon, D. Stello, and J. Yu, “Deep learning classification in asteroseis- mology,” Monthly Notices of the Royal Astronomical Society , vol. 469, no. 4, pp. 4578–4583, 2017

    2017

  42. [50]

    Uci machine learning repository,

    A. Asuncion and D. Newman, “Uci machine learning repository,” 2007

    2007

  43. [51]

    Columbia object image library (coil-20), 1996

    S. Nene, S. Nayar, H. Murase et al. , “Columbia object image library (coil-20), 1996.”

    1996

  44. [52]

    Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state,

    R. G. Andrzejak, K. Lehnertz, F. Mormann, C. Rieke, P. David, and C. E. Elger, “Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state,” Physical Review E , vol. 64, n...

    2001

  45. [53]

    Material perception: What can you see in a brief glance?

    L. Sharan, R. Rosenholtz, and E. Adelson, “Material perception: What can you see in a brief glance?” Journal of Vision , vol. 9, no. 8, pp. 784–784, 2009

    2009

  46. [54]

    Automated hate speech detection and the problem of offensive language,

    T. Davidson, D. Warmsley, M. Macy, and I. Weber, “Automated hate speech detection and the problem of offensive language,” in Proceedings of the International AAAI Conference on Web and Social Media, vol. 11, no. 1, 2017, pp. 512–515

    2017

  47. [55]

    Learning word vectors for sentiment analysis,

    A. Maas, R. E. Daly, P. T. Pham, D. Huang, A. Y . Ng, and C. Potts, “Learning word vectors for sentiment analysis,” in Proceedings of the 49th annual meeting of the association for computational linguistics: Human language technologies, 2011, pp. 142–150

    2011

  48. [56]

    “Kaggle,” https://www.kaggle.com

  49. [57]

    Classification of raisin grains using machine vision and artificial intelligence methods,

    ˙I. C ¸ INAR, M. KOKLU, and S ¸. TAS ¸DEM˙IR, “Classification of raisin grains using machine vision and artificial intelligence methods,” Gazi M¨uhendislik Bilimleri Dergisi (GMBD), vol. 6, no. 3, pp. 200–209, 2020

    2020

  50. [58]

    Application of rule induction algorithms for analysis of data collected by seismic hazard monitoring systems in coal mines,

    M. Sikora et al., “Application of rule induction algorithms for analysis of data collected by seismic hazard monitoring systems in coal mines,” Archives of Mining Sciences, vol. 55, no. 1, pp. 91–114, 2010

    2010

  51. [59]

    From group to individual labels using deep features,

    D. Kotzias, M. Denil, N. De Freitas, and P. Smyth, “From group to individual labels using deep features,” in Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining, 2015, pp. 597–606

    2015

  52. [60]

    Contributions to the study of sms spam filtering: new collection and results,

    T. A. Almeida, J. M. G. Hidalgo, and A. Yamakami, “Contributions to the study of sms spam filtering: new collection and results,” in Proceedings of the 11th ACM symposium on Document engineering , 2011, pp. 259– 262

    2011

  53. [61]

    A comparative user study of visualization techniques for cluster analysis of multidimensional data sets,

    E. Ventocilla and M. Riveiro, “A comparative user study of visualization techniques for cluster analysis of multidimensional data sets,” Informa- tion visualization, vol. 19, no. 4, pp. 318–338, 2020

    2020

  54. [62]

    Phishing detection based associative classification data mining,

    N. Abdelhamid, A. Ayesh, and F. Thabtah, “Phishing detection based associative classification data mining,”Expert Systems with Applications, vol. 41, no. 13, pp. 5948–5959, 2014

    2014

  55. [63]

    Investigating visual perception of degree centrality in graph visualization,

    X. Zhao, S. Fu, R. Yang, L. Yang, Y . Chen, J. Zhang, J. Long, F. Zhou, and Y . Zhao, “Investigating visual perception of degree centrality in graph visualization,” IEEE Transactions on Visualization and Computer Graphics, vol. 31, no. 6, pp. 3679–3692, 2025

    2025

  56. [64]

    Density peak clustering based on relative density relationship,

    J. Hou, A. Zhang, and N. Qi, “Density peak clustering based on relative density relationship,” Pattern Recognition, vol. 108, p. 107554, 2020. [Online]. Available: https://www.sciencedirect.com/science/article/pii/ S0031320320303575

    2020

  57. [65]

    Clustering by fast search and find of density peaks,

    A. Rodriguez and A. Laio, “Clustering by fast search and find of density peaks,” Science, vol. 344, no. 6191, pp. 1492–1496, 2014. [Online]. Available: https://www.science.org/doi/abs/10.1126/science.1242072

    2014 doi

  58. [66]

    Initialization is critical for preserving global data structure in both t-sne and umap,

    D. Kobak and G. C. Linderman, “Initialization is critical for preserving global data structure in both t-sne and umap,” Nature Biotechnology , vol. 39, no. 2, pp. 156–157, 2021

    2021

  59. [67]

    Dynamic programming,

    R. Bellman, “Dynamic programming,” Science, vol. 153, no. 3731, pp. 34–37, 1966

    1966

  60. [68]

    Efficient k-nearest neighbor graph construction for generic similarity measures,

    W. Dong, C. Moses, and K. Li, “Efficient k-nearest neighbor graph construction for generic similarity measures,” in Proceedings of the 20th International Conference on World Wide Web , ser. WWW ’11. New York, NY , USA: Association for Computing Machinery, 2011, p. 577–586

    2011

  61. [69]

    “normalized stress

    K. Smelser, J. Miller, and S. Kobourov, ““normalized stress” is not normalized: How to interpret stress correctly,” in 2024 IEEE Evaluation and Beyond - Methodological Approaches for Visualization (BELIV) , 2024, pp. 41–50

    2024

  62. [70]

    Nonlinear dimensionality reduction by locally linear embedding,

    S. T. Roweis and L. K. Saul, “Nonlinear dimensionality reduction by locally linear embedding,” Science, vol. 290, no. 5500, pp. 2323–2326, 2000

    2000

  63. [71]

    Scikit-learn: Machine learning in python,

    F. Pedregosa, G. Varoquaux, A. Gramfort, V . Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V . Dubourg et al. , “Scikit-learn: Machine learning in python,” the Journal of machine Learning research, vol. 12, pp. 2825–2830, 2011

    2011

  64. [72]

    Motta, “Lmds,” https://github.com/danilomotta/LMDS

    D. Motta, “Lmds,” https://github.com/danilomotta/LMDS

  65. [73]

    Zadu: A python library for evaluating the reliability of dimensionality reduction embeddings,

    H. Jeon, A. Cho, J. Jang, S. Lee, J. Hyun, H.-K. Ko, J. Jo, and J. Seo, “Zadu: A python library for evaluating the reliability of dimensionality reduction embeddings,” in 2023 IEEE Visualization and Visual Analytics (VIS), 2023, to appear

    2023

  66. [74]

    Geometric inference for probability measures,

    F. Chazal, D. Cohen-Steiner, and Q. M ´erigot, “Geometric inference for probability measures,” Foundations of Computational Mathematics , vol. 11, no. 6, pp. 733–751, 2011

    2011

  67. [75]

    Practical bayesian optimization of machine learning algorithms,

    J. Snoek, H. Larochelle, and R. Adams, “Practical bayesian optimization of machine learning algorithms,” in Advances in Neural Information Processing Systems, F. Pereira, C. Burges, L. Bottou, and K. Weinberger, Eds., vol. 25. Curran Associates, Inc., 2012

    2012

  68. [76]

    Quality assessment of dimensionality reduction: Rank-based criteria,

    J. A. Lee and M. Verleysen, “Quality assessment of dimensionality reduction: Rank-based criteria,” Neurocomputing, vol. 72, no. 7-9, pp. 1431–1443, 2009

    2009

  69. [77]

    Information retrieval perspective to nonlinear dimensionality reduction for data visu- alization

    J. Venna, J. Peltonen, K. Nybo, H. Aidos, and S. Kaski, “Information retrieval perspective to nonlinear dimensionality reduction for data visu- alization.” Journal of Machine Learning Research, vol. 11, no. 2, 2010

    2010

  70. [78]

    Understanding umap,

    A. Coenen and A. Pearce, “Understanding umap,” https://pair-code. github.io/understanding-umap/, 2019

    2019

  71. [79]

    Rcv1: A new bench- mark collection for text categorization research,

    D. D. Lewis, Y . Yang, T. Russell-Rose, and F. Li, “Rcv1: A new bench- mark collection for text categorization research,” Journal of machine learning research, vol. 5, no. Apr, pp. 361–397, 2004

    2004

  72. [80]

    Outlier detection: Methods, models, and classification,

    A. Boukerche, L. Zheng, and O. Alfandi, “Outlier detection: Methods, models, and classification,” ACM Comput. Surv. , vol. 53, no. 3, Jun

  73. [81]

    Ghostumap2: Measuring and analyzing (r,d)-stability of umap,

    M. Jung, T. Fujiwara, and J. Jo, “Ghostumap2: Measuring and analyzing (r,d)-stability of umap,” 2025. [Online]. Available: https://arxiv.org/abs/2507.17174

    2025 arXiv

  74. [82]

    Perception-based evaluation of projection methods for multidimensional data visualization,

    R. Etemadpour, R. Motta, J. G. d. S. Paiva, R. Minghim, M. C. F. de Oliveira, and L. Linsen, “Perception-based evaluation of projection methods for multidimensional data visualization,” IEEE Transactions on Visualization and Computer Graphics, vol. 21, no. 1, pp. 81–94, 2015

    2015

  75. [83]

    Winglets: Visualizing association with uncertainty in multi-class scat- terplots,

    M. Lu, S. Wang, J. Lanir, N. Fish, Y . Yue, D. Cohen-Or, and H. Huang, “Winglets: Visualizing association with uncertainty in multi-class scat- terplots,” IEEE Transactions on Visualization and Computer Graphics , vol. 26, no. 1, pp. 770–779, 2020

    2020

  76. [84]

    A perception-driven approach to supervised dimensionality reduction for visualization,

    Y . Wang, K. Feng, X. Chu, J. Zhang, C.-W. Fu, M. Sedlmair, X. Yu, and B. Chen, “A perception-driven approach to supervised dimensionality reduction for visualization,” IEEE Transactions on Visualization and Computer Graphics, vol. 24, no. 5, pp. 1828–1840, 2018

    2018

  77. [85]

    H. Xiao, K. Rasul, and R. V ollgraf. (2017) Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms

    2017

  78. [86]

    How to use t-sne effectively,

    M. Wattenberg, F. Vi ´egas, and I. Johnson, “How to use t-sne effectively,” Distill, 2016. [Online]. Available: http://distill.pub/2016/misread-tsne

    2016

  79. [87]

    Automatic selection of t-sne perplexity,

    Y . Cao and L. Wang, “Automatic selection of t-sne perplexity,” arXiv preprint arXiv:1708.03229, 2017

    2017 arXiv

  80. [88]

    Gpgpu linear complexity t- sne optimization,

    N. Pezzotti, J. Thijssen, A. Mordvintsev, T. H ¨ollt, B. Van Lew, B. P. Lelieveldt, E. Eisemann, and A. Vilanova, “Gpgpu linear complexity t- sne optimization,” IEEE Transactions on Visualization and Computer Graphics, vol. 26, no. 1, pp. 1172–1181, 2019. IEEE TRANSACTIONS ON ...

    2019

  81. [89]

    Bringing umap closer to the speed of light with gpu acceleration,

    C. J. Nolet, V . Lafargue, E. Raff, T. Nanditale, T. Oates, J. Zedlewski, and J. Patterson, “Bringing umap closer to the speed of light with gpu acceleration,” arXiv preprint arXiv:2008.00325, 2020

    2008 arXiv

  82. [90]

    Fpga implemen- tation of the principal component analysis algorithm for dimensionality reduction of hyperspectral images,

    D. Fernandez, C. Gonzalez, D. Mozos, and S. Lopez, “Fpga implemen- tation of the principal component analysis algorithm for dimensionality reduction of hyperspectral images,”Journal of Real-Time Image Process- ing, vol. 16, pp. 1395–1406, 2019

    2019

  83. [91]

    Progressive uniform manifold approximation and projection,

    H.-K. Ko, J. Jo, and J. Seo, “Progressive uniform manifold approximation and projection,” in 22nd Eurographics Conference on Visualization, EuroVis 2020-Short Papers. Eurographics Association, 2020, pp. 133– 137

    2020

  84. [92]

    Parametric UMAP Embeddings for Representation and Semisupervised Learning,

    T. Sainburg, L. McInnes, and T. Q. Gentner, “Parametric UMAP Embeddings for Representation and Semisupervised Learning,” Neural Computation, vol. 33, no. 11, pp. 2881–2907, 10 2021

    2021

  85. [93]

    Clustme: A visual quality measure for ranking monochrome scatterplots based on cluster patterns,

    M. M. Abbas, M. Aupetit, M. Sedlmair, and H. Bensmail, “Clustme: A visual quality measure for ranking monochrome scatterplots based on cluster patterns,” Computer Graphics Forum, vol. 38, no. 3, pp. 225–236, 2019

    2019

  86. [94]

    Predicting user preferences of dimensionality reduction embedding quality,

    C. Morariu, A. Bibal, R. Cutura, B. Fr ´enay, and M. Sedlmair, “Predicting user preferences of dimensionality reduction embedding quality,” IEEE Transactions on Visualization and Computer Graphics, vol. 29, no. 1, pp. 745–755, 2023

    2023

  87. [95]

    Understanding bias in perceiving dimensionality reduction projections,

    S. Doh, H. Jeon, S. Shin, G. J. Quadri, N. W. Kim, and J. Seo, “Understanding bias in perceiving dimensionality reduction projections,” arXiv preprint arXiv:2507.20805, 2025

    2025 arXiv

  88. [96]

    Efficient, high-quality force-directed graph drawing,

    Y . Hu, “Efficient, high-quality force-directed graph drawing,”Mathemat- ica journal, vol. 10, no. 1, pp. 37–71, 2005

    2005

  89. [97]

    Graph layouts by t-sne,

    J. F. Kruiger, P. E. Rauber, R. M. Martins, A. Kerren, S. Kobourov, and A. C. Telea, “Graph layouts by t-sne,” Computer Graphics Forum, vol. 36, no. 3, pp. 283–294, 2017. Hyeon Jeon is a Ph.D. Student at the Depart- ment of Computer Science and Engineering, Seoul National Univ...

    2017

  90. [2020]

    Available: https://doi.org/10.1145/3381028

    [Online]. Available: https://doi.org/10.1145/3381028

    doi

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.