Pith. sign in

REVIEW 3 major objections 4 minor 43 references

TemporalFlowViz: Parameter-Aware Visual Analytics for Interpreting Scramjet Combustion Evolution

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read TemporalFlowViz argues that latent-space trajectories of ViT-embedded flow fields reveal scramjet combustion modes and evolution patterns across hundreds of simulations.

desk verdict Useful visual analytics pipeline for scramjet simulations, but the main effectiveness claim rests on unvalidated ViT embeddings and qualitative expert feedback. read the letter →

arxiv 2509.04834 v1 pith:AY4OJQBI submitted 2025-09-05 cs.CV

Yifei Jia , Shiyu Cheng , Yu Dong , Guan Li , Dong Tian , Ruixiao Peng , Xuyi Lu , Yu Wang
show 2 more authors
Wei Yao Guihua Shan
This is my paper · ORCID
classification cs.CV
keywords scramjetcombustionvisualanalyticstemporalflowfieldanalysisVisionTransformerembeddingUMAPDBSCANtrajectoryvision-languagesummarization
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

TemporalFlowViz sets out to show that the evolution of a scramjet combustion simulation can be read from the geometry of its flow-field images. The paper argues that cropping each frame to the isolator–cavity region, embedding the cropped pressure and OH fields with pretrained Vision Transformers, and projecting those embeddings with UMAP, a dimension-reduction method chosen for its preservation of trajectory shape, produces clusters that correspond to physically meaningful combustion modes. Chronologically connecting each case's projected frames yields trajectories whose shape—converging, oscillating, or diverging—reveals whether combustion stabilizes, transitions, or fails. Expert annotations on cluster centroids then prompt a vision-language model to write frame-level and case-level summaries, giving experts a reusable semantic layer across more than 200 simulations and approximately three terabytes of data. The payoff would be that hypothesis generation and cross-case comparison in high-speed propulsion analysis no longer require frame-by-frame manual inspection.

What carries the argument

The carrying object is the temporal trajectory in a two-dimensional latent space. Each cropped pressure or OH frame $P'_i(t)$ is embedded as $p_i(t)=\mathrm{ViT}(P'_i(t),W_P)$, projected by UMAP to $p'_i(t)\in\mathbb{R}^2$, and assigned to a cluster by the density-based algorithm DBSCAN; the frame points of a case are connected in time to form its trajectory. Cluster centroids are found by taking the embedding nearest the cluster mean, and experts annotate these centroids; the annotations plus nearest-centroid images condition a vision-language model that writes frame and case summaries. Trajectory similarity uses a time-warped distance that normalizes frame differences by local motion magnitudes, so cases with similar late-stage convergence can be retrieved despite minor timing shifts. Region cropping around the isolator–cavity zone is a preprocessing step the paper shows to reduce trajectory convergence radius by about 35 percent.

What would settle it

Take the exact cases featured in Section 8, ask two combustion experts to label each frame's mode independently, and compare those labels with the DBSCAN cluster memberships under the paper's default settings. If same-mode frames spread across distinct clusters or different-mode frames share one cluster on more than a small fraction of frames, the claimed correspondence between latent clusters and combustion modes fails; the paper offers only visual agreement so far.

Watch

Extended reading notes

Core claim

The paper's central claim is that latent embeddings of cropped scramjet flow-field frames carry combustion-relevant structure that standard scalar metrics leave hidden. Concretely, it claims that the clusters produced by the density-based algorithm DBSCAN on UMAP-projected ViT embeddings separate distinct combustion modes—in the case studies, ramjet-like pressure behavior, stable scramjet operation, shear-layer flame combustion, and jet-wake combustion—and that the chronological path of a case through this latent space is a reliable signature of its evolution: stable cases stay in compact clusters, transitional cases move between clusters, and similar trajectories correspond to physically similar cases under different initial conditions. It further claims that expert-written labels on selected cluster centroids can condition a vision-language model to produce accurate natural-language descriptions of individual frames and whole cases, turning the latent structure into a shareable interpretive layer.

Load-bearing premise

The pipeline works only if the visual features extracted from the flow-field images group together exactly the frames that a combustion expert would call the same mode; if the clusters reflect image artifacts rather than combustion physics, the trajectories and reports mislead.

Editorial extensions

If this is right

  • With this workflow, an expert can classify thousands of frames into combustion modes without inspecting every frame; the case studies identify shear-layer and jet-wake modes from OH-field trajectories.
  • Trajectory shape becomes a diagnostic: compact trajectories indicate stable scramjet combustion, trajectories leaving a compact cluster indicate ramjet-like transition, and fluctuations that stay within one cluster indicate a stable mode with flame-shape variability.
  • Similar-trajectory retrieval lets experts generalize a finding from one case to other cases with different initial conditions, supporting parameter-aware comparison of hundreds of simulations.
  • A single expert annotation on a cluster centroid propagates to many frames and cases through vision-language summarization, so one annotation can support dozens of frames rather than requiring per-frame effort.

Reading between the lines

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

  • Editorial inference: the paper's crop-and-embed choice could be tested quantitatively by measuring cluster purity against expert labels on a held-out frame sample; the current support is visual agreement, not a reported number.
  • Editorial inference: because the three vision encoders produce different latent structures, the downstream clusters are not invariant to backbone choice; a practitioner would need a selection criterion such as convergence radius or label agreement rather than visual inspection.
  • Editorial inference: the measure of how tightly trajectories settle, which the paper uses to justify cropping, could itself serve as a scalar stability indicator for triaging hundreds of simulations, independent of the visual interface.
  • Editorial inference: the same pipeline likely transfers to other unsteady propulsion simulation outputs, but only after re-testing the embedding assumption; the paper demonstrates scramjet pressure and OH fields only.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper presents TemporalFlowViz, a visual analytics workflow and system for exploring and interpreting temporal flow-field data from scramjet combustion simulations. The pipeline encodes cropped flow-field images (pressure and OH fields) with pretrained Vision Transformers, projects the embeddings with UMAP, clusters them with DBSCAN, and constructs temporal trajectories in the latent space. Domain experts annotate cluster centroids, and these annotations are used to condition a vision-language model (Gemma-3) that generates frame-level and case-level textual summaries. The system also provides parameter filtering, similarity-based trajectory retrieval, and coordinated multi-view exploration. The authors evaluate the system through two expert-informed case studies and qualitative expert feedback, claiming that TemporalFlowViz enhances hypothesis generation, supports interpretable pattern discovery, and improves knowledge discovery in large-scale scramjet combustion analysis.

Significance. If the central claims are accepted, TemporalFlowViz would be a useful contribution to scientific visual analytics, demonstrating a workflow that combines deep pretrained embeddings, density-based clustering, and expert-grounded vision-language summarization for a domain-specific simulation dataset. The paper's strengths include a clear articulation of domain requirements from expert collaboration, an explicit discussion of limitations in Section 9.2, and the release of code and a video demonstration. However, the significance is conditional on whether the latent clusters and trajectories correspond to physically meaningful combustion modes and on whether the case studies provide convincing evidence of effectiveness; the current qualitative evaluation does not fully establish these points.

major comments (3)
  1. [Sections 6.2, 6.3, 6.5, and 8] The central claim that TemporalFlowViz uncovers latent combustion modes and enables interpretable pattern discovery rests on the assumption that ViT embeddings, UMAP projections, and DBSCAN clusters isolate physically meaningful combustion states. Section 6.2 selects among ViT backbones by visual inspection of trajectory continuity and cluster separability, Section 6.3 selects UMAP over t-SNE by visual evaluation and expert feedback, and Section 8.2 describes expert E4 interactively adjusting the DBSCAN eps and minSamples during the case study. Section 9.2 concedes that the ViTs are general-purpose and that task-specific fine-tuning is future work. Yet no independent quantitative validation ties the resulting cluster assignments to physical ground truth (for example, expert-defined mode labels, pressure oscillation metrics, or ignition delay). Without such validation, the identified clusters and trajectory-based summaries could reflect incidental image features such as colormap statistics, background cropping, or global brightness rather than combustion physics, and the retrieval, reports, and conclusions would be systematically misleading.
  2. [Section 6.4] The temporal dissimilarity metric is not presented in a usable form. The displayed equation is garbled: the expression 'pi(a)-p j(b)' with misplaced parentheses and incomplete norm notation cannot be parsed, and the claimed time-warped alignment is not clearly defined. Because this metric is used to compute the top-k similar trajectories in the Similar Trajectories View and to support cross-case comparison, the manuscript should provide a clean, unambiguous definition (for instance, with proper vector norms and index notation) and ideally a simple numerical example to demonstrate its behavior. As written, the method cannot be implemented or checked.
  3. [Sections 8 and 10] The abstract and conclusion state that TemporalFlowViz enhances hypothesis generation and knowledge discovery, but the supporting evidence consists of two qualitative case studies performed by invited experts E3 and E4 plus post-hoc verbal feedback. There is no pre-registered study protocol, no inter-rater reliability assessment, no quantified task performance, and no comparison to a baseline tool or alternative approach. Moreover, because the experts interactively tuned clustering parameters during the case studies, the reported findings are vulnerable to confirmation bias. The authors should either temper the effectiveness claims to match the qualitative evidence or add a more controlled evaluation (for example, a comparative study with a baseline system and fixed default parameters, or a quantitative linking of cluster labels to physical diagnostics).
minor comments (4)
  1. [Throughout] There are inconsistencies in capitalization and model naming, for example 'BLIP2-opt-2.7b' in Section 6.2 versus 'Blip2-opt-2.7b' in Section 8.1, and 'dbscan' in the Filtering Panel description versus 'DBSCAN' elsewhere; these should be made consistent.
  2. [Section 6.1] The optional spatial weighting matrix W_P is introduced but never explained or defined; the manuscript should clarify how it is constructed, whether it is applied to images before embedding or to the embeddings, and how its values were chosen for the case studies.
  3. [Section 6.5 and Figure 8] The text states that DBSCAN is applied to UMAP-projected 2D embeddings, but it is not explicitly stated whether the clustering is performed separately for each field (pressure and OH) or jointly; Figure 8 suggests separate analyses, but the caption and text should be explicit to avoid ambiguity.
  4. [Section 8.1] The phrase 'We drawed dashed circles' should be corrected to 'We drew dashed circles' or a more formal phrasing, and figure callouts should be checked for consistency (e.g., '(a1)' is used both for a scatter plot and a zoomed-in view).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is a descriptive visual analytics system; its central claims rest on expert feedback and case studies, not on a derivation that reduces to its inputs.

full rationale

TemporalFlowViz does not contain a predictive derivation chain. The pipeline uses external pretrained Vision Transformers, standard UMAP projection, and DBSCAN clustering; these components are not fitted to the claimed outcomes such as 'hypothesis generation' or 'knowledge discovery.' The only place where expert labels enter the output is the semantic summarization stage, where the paper explicitly states that expert annotations are used as context for the VLM (Section 6.5 and Section 7.5). This means the summaries are designed to reflect expert terminology by construction, and the paper presents this as the mechanism rather than as an independent discovery, so it is not a hidden circularity. The quantitative preprocessing comparison (Section 6.1) uses an independently defined convergence radius and compares cropped versus whole frames through the same embedding pipeline, which is a legitimate empirical comparison rather than a self-referential fit. The two cited works by overlapping authors (refs [24] and [40]) appear only in related-work enumerations and are not load-bearing for any claim. The main weakness identified by the skeptic, that generic ViT embeddings may not separate physically meaningful combustion modes, is a question of external validity or grounding, not circularity, because the paper makes no claim that the clusters are derived from first principles. The evaluation is qualitative expert feedback and case studies, which is weak evidence but not circular. Therefore the appropriate finding is no significant circularity with score 0.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The system relies on several hand-chosen parameters and domain assumptions rather than a derivation from first principles. The free parameters are design choices, and the axioms are unverified premises about embedding quality and annotation validity.

free parameters (8)
  • UMAP n_neighbors = 15
    Chosen per Section 6.3 following recommended values and empirical tuning.
  • UMAP min_dist = 0.1
    Set per Section 6.3 as part of the fixed UMAP configuration.
  • DBSCAN eps = user-set
    Clustering parameter adjusted interactively by users in the system, affecting cluster boundaries.
  • DBSCAN minSamples = user-set
    Clustering parameter adjusted interactively, affecting noise and cluster size.
  • Crop box location and size = fixed around isolator and cavity
    Chosen based on expert input and validated by convergence radius comparison in Section 6.1.
  • Spatial weighting matrix W_P = hand-assigned
    Optional weights assigned to subregions based on domain knowledge in the case studies.
  • Number of nearest centroids k for VLM context = unspecified
    Used for retrieving context images and annotations, value not reported.
  • K=5 for convergence radius = 5
    Used in Section 6.1 to define trajectory convergence radius, chosen without justification.
assumptions (5)
  • domain assumption Pretrained Vision Transformer embeddings capture combustion-relevant visual features in flow-field images.
    Core to the pipeline; the paper provides no quantitative validation that these embeddings separate combustion modes better than other features.
  • domain assumption UMAP projection preserves temporal trajectory shapes and cluster structure.
    Needed for trajectory reasoning; selected by visual comparison with t-SNE, not by a quantitative metric.
  • domain assumption DBSCAN clusters correspond to distinct latent combustion modes.
    The interpretation of clusters as combustion modes is assumed and used throughout the case studies.
  • domain assumption Expert annotation of cluster centroids provides accurate semantic labels for vision-language grounding.
    The vision-language summaries inherit correctness from these annotations, but annotation accuracy is not measured.
  • domain assumption Vision-language model summaries accurately describe frames based on nearest centroid context.
    No evaluation of summary correctness or consistency is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TemporalFlowViz: Parameter-Aware Visual Analytics for Interpreting Scramjet Combustion Evolution." pith.science (2026). https://pith.science/paper/AY4OJQBI

@misc{pith2026250904834,
  author       = {Pith},
  title        = {Pith review of: TemporalFlowViz: Parameter-Aware Visual Analytics for Interpreting Scramjet Combustion Evolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AY4OJQBI}},
  note         = {Machine review of arXiv:2509.04834}
}
read the original abstract

Understanding the complex combustion dynamics within scramjet engines is critical for advancing high-speed propulsion technologies. However, the large scale and high dimensionality of simulation-generated temporal flow field data present significant challenges for visual interpretation, feature differentiation, and cross-case comparison. In this paper, we present TemporalFlowViz, a parameter-aware visual analytics workflow and system designed to support expert-driven clustering, visualization, and interpretation of temporal flow fields from scramjet combustion simulations. Our approach leverages hundreds of simulated combustion cases with varying initial conditions, each producing time-sequenced flow field images. We use pretrained Vision Transformers to extract high-dimensional embeddings from these frames, apply dimensionality reduction and density-based clustering to uncover latent combustion modes, and construct temporal trajectories in the embedding space to track the evolution of each simulation over time. To bridge the gap between latent representations and expert reasoning, domain specialists annotate representative cluster centroids with descriptive labels. These annotations are used as contextual prompts for a vision-language model, which generates natural-language summaries for individual frames and full simulation cases. The system also supports parameter-based filtering, similarity-based case retrieval, and coordinated multi-view exploration to facilitate in-depth analysis. We demonstrate the effectiveness of TemporalFlowViz through two expert-informed case studies and expert feedback, showing TemporalFlowViz enhances hypothesis generation, supports interpretable pattern discovery, and enhances knowledge discovery in large-scale scramjet combustion analysis.

Figures

Figures reproduced from arXiv: 2509.04834 by the authors.

Figure 1
Figure 1. The visual analytics system interface of TemporalFlowViz: (a) the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 3
Figure 3. Comparison of UMAP-projected trajectories using pressure [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. UMAP versus t-SNE: The top two scatter plots display the dimensionality reduction results of the pressure and OH fields’ em￾beddings obtained through t-SNE, while the bottom two plots present the corresponding results using UMAP. UMAP better preserves tra￾jectory shapes and cluster separations at OH fields compared to t￾SNE. UMAP-projected frame embeddings as: p ′ i (1), p ′ i (2), ..., p ′ i (ti). These points are … view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: The case study 1: Ramjet Pressure field exploration. Cropped pressure field frames were processed by InternViT and clustered using DBSCAN. Scatter plot (a1) exhibited the clustering result in the Temporal Trajectory View. (b1) and (c1) were the 2 most similar trajector…
Figure 6
Figure 6. Figure 6: The case study 1: Scramjet Pressure fields Explo￾ration. The temporal trajectory of selected case resided in a small cluster (a), zoomed in to (a1), its flow field frames were in figure (af), and similar trajectories at (a2) and (a3). Clusters (a) to (d) had scramjet c…
Figure 8
Figure 8. Figure 8: Case Study 2: Clustering of OH Field Trajectories to Identify Combustion Modes. The temporal trajectory of the selected case was shown in region (s1), with its two most similar trajectories located at (s2) and (s3). The corresponding OH field images for these cases wer…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 35 canonical work pages

  1. [1]

    J. Bae, T. Helldin, M. Riveiro, S. Nowaczyk, M.-R. Bouguelia, and G. Falkman. Interactive clustering: A comprehensive review.ACM Comput. Surv., 53(1), Feb. 2020. doi: 10.1145/3340960 2

    work page doi:10.1145/3340960 2020

  2. [2]

    Bostock, V

    M. Bostock, V . Ogievetsky, and J. Heer. D3 data-driven docu- ments.IEEE Transactions on Visualization and Computer Graphics, 17(12):2301–2309, Dec. 2011. doi: 10.1109/TVCG.2011.185 2

    doi:10.1109/tvcg.2011.185 2011

  3. [3]

    D. Cai, X. He, X. Wang, H. Bao, and J. Han. Locality preserving non- negative matrix factorization. InProceedings of the 21st International Joint Conference on Artificial Intelligence, IJCAI’09, p. 1010–1015. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 2009. 2

    work page 2009

  4. [4]

    H. Chen, S. Jiang, X. Yu, H. Yin, X. Wang, Y . Hu, C. Wang, and C. Li. Geovis: a data-driven geographic visualization recommendation sys- tem via latent space encoding.Journal of Visualization, 27(4):603– 622, 2024. 2

    work page 2024

  5. [5]

    J. Chen, H. Huang, H. Ye, Z. Peng, C. Li, and C. Wang. Salientime: user-driven selection of salient time steps for large-scale geospatial data visualization. InProceedings of the 2024 CHI conference on human factors in computing systems, pp. 1–19, 2024. 2

    work page 2024

  6. [6]

    Y . Chen, Y . Zhao, X. Li, J. Zhang, J. Long, and F. Zhou. An open dataset of data lineage graphs for data governance research.Visual Informatics, 8(1):1–5, 2024. doi: 10.1016/j.visinf.2024.01.001 2

    work page doi:10.1016/j.visinf.2024.01.001 2024

  7. [7]

    Z. Chen, J. Wu, W. Wang, W. Su, G. Chen, S. Xing, M. Zhong, Q. Zhang, X. Zhu, L. Lu, et al. Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 24185–24198, 2024. 5

    work page 2024

  8. [8]

    Chidananda Gowda and G

    K. Chidananda Gowda and G. Krishna. Agglomerative clustering us- ing the concept of mutual nearest neighbourhood.Pattern Recogni- tion, 10(2):105–112, 1978. doi: 10.1016/0031-3203(78)90018-3 2

    doi:10.1016/0031-3203(78)90018-3 1978

Show all 43 references
  1. [9]

    Curran, W

    E. Curran, W. Heiser, and D. Pratt. Fluid phenomena in scramjet com- bustion systems.Annual Review of Fluid Mechanics, 28:323–360, 11

  2. [10]

    Dalal and B

    N. Dalal and B. Triggs. Histograms of oriented gradients for hu- man detection. In2005 IEEE Computer Society Conference on Com- puter Vision and Pattern Recognition (CVPR’05), vol. 1, pp. 886–893. IEEE, 2005. 2

    2005

  3. [11]

    Dosovitskiy, L

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, and Others. An image is worth 16x16 words: Transformers for image recognition at scale. InInternational Conference on Learning Repre- sentations, 2021. 2

    2021

  4. [12]

    Eirich, J

    J. Eirich, J. Bonart, D. J ¨ackle, M. Sedlmair, U. Schmid, K. Fischbach, T. Schreck, and J. Bernard. Irvine: A design study on analyzing cor- relation patterns of electrical engines.IEEE Transactions on Visual- ization and Computer Graphics, 28(1):11–21, 2021. 2

    2021

  5. [13]

    Ester, H.-P

    M. Ester, H.-P. Kriegel, J. Sander, and X. Xu. A density-based al- gorithm for discovering clusters in large spatial databases with noise. InProceedings of the Second International Conference on Knowledge Discovery and Data Mining (KDD), pp. 226–231, 1996. 2

    1996

  6. [14]

    E. Fini, M. Shukor, X. Li, P. Dufter, M. Klein, D. Haldimann, S. Aitharaju, V . G. T. da Costa, L. B ´ethune, Z. Gan, A. T. Toshev, M. Eichner, M. Nabi, Y . Yang, J. M. Susskind, and A. El-Nouby. Mul- timodal autoregressive pre-training of large vision encoders, 2024. 5

    2024

  7. [15]

    K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition.Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016. 2

    2016

  8. [16]

    Z. He, C. Chen, Y . Wu, X. Tian, Q. Chu, Z. Huang, and W. Zhang. Real-time interactive parallel visualization of large-scale flow-field data.Applied Sciences, 13(16):9092, 2023. doi: 10.3390/ app13169092 2

    2023

  9. [17]

    Huang, J

    D. Huang, J. Lai, and C.-D. Wang. Ensemble clustering using factor graph.Pattern Recognition, 50:131–142, 2016. doi: 10.1016/j.patcog .2015.08.015 2

    2016 doi

  10. [18]

    Krizhevsky, I

    A. Krizhevsky, I. Sutskever, and G. E. Hinton. Imagenet classifica- tion with deep convolutional neural networks. InAdvances in neural information processing systems, pp. 1097–1105, 2012. 2

    2012

  11. [19]

    Ladeinde.A Critical Review of Scramjet Combustion Simulation (Invited)

    F. Ladeinde.A Critical Review of Scramjet Combustion Simulation (Invited). 2009. doi: 10.2514/6.2009-127 1

    2009 doi

  12. [20]

    C. Y . Li, X. Liang, Z. Hu, and E. P. Xing. Knowledge-driven encode, retrieve, paraphrase for medical image report generation.Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):6666–6673, Jul. 2019. doi: 10.1609/aaai.v33i01.33016666 2

    2019 doi

  13. [21]

    J. Li, D. Li, S. Savarese, and S. Hoi. Blip-2: Bootstrapping language- image pre-training with frozen image encoders and large language models, 2023. 5

    2023

  14. [22]

    D. G. Lowe. Distinctive image features from scale-invariant key- points.International journal of computer vision, 60(2):91–110, 2004. 2

    2004

  15. [23]

    McInnes, J

    L. McInnes, J. Healy, and J. Melville. Umap: Uniform manifold ap- proximation and projection for dimension reduction, 2020. 2

    2020

  16. [24]

    R. Peng, Y . Dong, G. Li, D. Tian, and G. Shan. Textlens: large language models-powered visual analytics enhancing text clustering. Journal of Visualization, pp. 1–19, 2025. 2

    2025

  17. [25]

    Y . A. Reddy, J. Wahl, and M. Sj¨odahl. Twins-pivnet: Spatial attention- based deep learning framework for particle image velocimetry using vision transformer.Ocean Engineering, 318:120205, 2025. doi: 10. 1016/j.oceaneng.2024.120205 2

    2025

  18. [26]

    W. Shi, Y . Tian, J. Le, and F. Zhong. Effect of pilot hydrogen on the formation of dynamic flame in an ethylene-fueled scramjet with a cavity.Physics of Fluids, 33(5), 2021. 7

    2021

  19. [27]

    G. Team, A. Kamath, J. Ferret, S. Pathak, N. Vieillard, R. Merhej, S. Perrin, T. Matejovicova, A. Ram ´e, M. Rivi `ere, et al. Gemma 3 technical report.arXiv preprint arXiv:2503.19786, 2025. 6

    2025 arXiv

  20. [28]

    van der Maaten and G

    L. van der Maaten and G. Hinton. Visualizing data using t-sne.Jour- nal of Machine Learning Research, 9(86):2579–2605, 2008. 2

    2008

  21. [29]

    Vincent, H

    P. Vincent, H. Larochelle, Y . Bengio, and P.-A. Manzagol. Stacked de- noising autoencoders: Learning useful representations in a deep net- work with a local denoising criterion.Journal of machine learning research, 11:3371–3408, 2010. 2

    2010

  22. [30]

    J. Wang, J. Wang, Q. Ke, G. Zeng, and S. Li. Fast approximate k- means via cluster closures. In2012 IEEE Conference on Computer Vision and Pattern Recognition, pp. 3037–3044, 2012. doi: 10.1109/ CVPR.2012.6248034 2

    2012

  23. [31]

    J. Wang, J. Wang, J. Song, X.-S. Xu, H. T. Shen, and S. Li. Optimized cartesian k-means.IEEE Transactions on Knowledge and Data En- gineering, 27(1):180–192, 2015. doi: 10.1109/TKDE.2014.2324592 2

    2015

  24. [32]

    M. Wang, J. Tao, J. Ma, Y . Shen, and C. Wang. Flowvisual: A visual- ization app for teaching and understanding 3d flow field concepts. In Proceedings of the IEEE Pacific Visualization Symposium, pp. 133–

  25. [33]

    S. Wang, J. H. Seidman, S. Sankaran, H. Wang, G. J. Pappas, and P. Perdikaris. Cvit: Continuous vision transformer for operator learn- ing, 2025. 2

    2025

  26. [34]

    S. Wang, Z. Zhao, X. Ouyang, Q. Wang, and D. Shen. Chatcad: In- teractive computer-aided diagnosis on medical image using large lan- guage models, 2023. 3

    2023

  27. [35]

    Wang, Y .-L

    W.-T. Wang, Y .-L. Wu, C.-Y . Tang, and M.-K. Hor. Adaptive density- based spatial clustering of applications with noise (dbscan) accord- ing to data. In2015 International Conference on Machine Learning and Cybernetics (ICMLC), vol. 1, pp. 445–451, 2015. doi: 10.1109/ ICMLC.20...

    2015

  28. [36]

    Z. Wang, L. Liu, L. Wang, and L. Zhou. R2gengpt: Radiology report generation with frozen llms, 2023. 3

    2023

  29. [37]

    C. Wu, Y . Chen, Y . Dong, F. Zhou, Y . Zhao, and C. J. Liang. Vizoptics: Getting insights into optics via interactive visual analysis.Computers and Electrical Engineering, 107:108624, 2023. 2

    2023

  30. [38]

    D. Yu, O. Ian, L. Jie, Y . Xiaoru, and N. Q. Vinh. User-centered visual explorer of in-process comparison in spatiotemporal space.Journal of Visualization, 26(2):403–421, 2023. 2

    2023

  31. [39]

    Zelnik-Manor and P

    L. Zelnik-Manor and P. Perona. Self-tuning spectral clustering. In Neural Information Processing Systems, 2004. 2

    2004

  32. [40]

    Zhang, Y

    C. Zhang, Y . Dong, Y . Wang, Y . Han, G. Shan, and B. Tang. Auragenome: An llm-powered framework for on-the-fly reusable and scalable circular genome visualizations.IEEE Computer Graphics and Applications, (01):1–14, 2025. 2

    2025

  33. [41]

    X. Zhao, X. Wang, X. Zou, H. Liang, G. Bai, N. Zhang, X. Huang, F. Zhou, and Y . Zhao. Graph visualization efficiency of popular web- based libraries.Visual Computing for Industry, Biomedicine, and Art, 8(1):12, May 2025. doi: 10.1186/s42492-025-00193-y 2

    2025 doi

  34. [42]

    Znamenskaya

    I. Znamenskaya. Methods for panoramic visualization and digital analysis of thermophysical flow fields: A review.arXiv preprint arXiv:2108.01464, 2021. 2

    2021 arXiv

  35. [2003]

    doi: 10.1146/annurev.fl.28.010196.001543 1

Pith tools

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