Pith. sign in

REVIEW 4 major objections 5 minor 55 references

Topology Guidance: Controlling the Outputs of Generative Models via Vector Field Topology

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read A training-free guidance method steers a diffusion model over vector fields so that generated fields contain user-specified critical points—sinks, sources, and saddles at chosen locations—while staying in the modeled data distribution.

desk verdict Promising idea for topology-guided diffusion, but the stability/focus guidance as written is mathematically inconsistent with the reported results—fix before trusting the numbers. read the letter →

arxiv 2505.06804 v1 pith:IN7EAK3N submitted 2025-05-11 cs.LG stat.ML

classification cs.LGstat.ML
keywords topologyguidancevectorfieldcriticalpointsdiffusioncoordinate-basedneuralnetworkssimulationensemblesgenerativecontrolfluidflowanalysis
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a diffusion model trained on latent codes of vector fields can be steered, without any retraining, to produce fields with user-specified critical points: the right location, the right type (sink, source, or saddle), and the right stability (node or focus). The proposed method, topology guidance, turns the differentiable structure of a coordinate-based neural field into energy functions, then adds gradients of those energies to the diffusion model's denoising step. On 2D fluid-flow ensembles, the paper reports alignment rates above 85% and often near 100% for fixed noise seeds, while Fréchet distances to the real ensemble remain close to unguided sampling. If correct, this means an analyst can query a generative model by topological description instead of drawing many samples and checking them afterward.

What carries the argument

The central object is the field-conditioned coordinate network f_phi(p; z_hat_t), mapping a spatial location p and a latent code z into a 2D vector and differentiable in both arguments. It supplies three signals that define the guidance energies: the field norm ||f_phi(p; z_hat_t)|| for critical-point existence, the eigenvalues of the Jacobian J_phi(p; z_hat_t) for sink/source/saddle type, and the discriminant $\Delta$ = Tr(J)^2 - 4|J| for node versus focus stability. Sigmoid penalties push eigenvalue signs and $\Delta$ toward the user-specified classification, and all penalties are added together for multiple critical points. The diffusion model provides the latent prior, while DDIM's predicted clean latent z_hat_t is what keeps the coordinate network's Jacobian meaningful at intermediate timesteps.

What would settle it

A decisive test is to sweep the timestep at which guidance is first applied, holding one noise seed fixed, and record critical-point alignment: the paper's account predicts a plateau above 85% for a range of start times, and absence of that plateau, or alignment that only appears when the critical-point detector's distance tolerance is loosened, would refute the central claim. A complementary numerical check computes the distance between DDIM-predicted clean latents at the chosen start time and the training latent distribution; a large distance would violate the in-distribution assumption stated in Sec. 4.1.

Watch

Extended reading notes

Core claim

Topology guidance claims that a diffusion model over latent field codes can be guided to satisfy a prescribed critical-point configuration by evaluating a coordinate-based neural network with sinusoidal activations on the DDIM-predicted clean latent at the specified location. The network supplies the field value, its Jacobian, and the trace-determinant discriminant, which are converted into energy terms for critical-point existence, type, and stability. Gradients of the summed energy with respect to the noisy latent are added to the model's predicted noise, steering each denoising step toward the desired topology. The paper reports alignment above 85% across single and combined specifications, higher alignment when the noise vector is fixed, and Fréchet distances comparable to unguided generation, concluding that the generated fields satisfy the specified topology while remaining in the data distribution.

Load-bearing premise

The method assumes that the clean latent predicted by DDIM at the timesteps where guidance is active is close enough to the latent distribution the coordinate network was trained on for its Jacobian-based energy gradients to be meaningful; the paper states that raw noisy latents are out-of-distribution at large t and provides no quantitative check that the predicted latent fixes this.

Editorial extensions

If this is right

  • A pretrained diffusion model of fields can be queried by topological specification without retraining or fine-tuning, so users can synthesize many fields sharing a prescribed critical-point configuration.
  • Topology guidance enables direct comparison of two simulation ensembles: fields from different distributions can be generated with identical prescribed topology, exposing differences in local flow behavior.
  • The guidance is largely local, so moving the prescribed critical point changes the field mainly in a neighborhood of that point, supporting what-if edits of individual samples.
  • Combining norm, eigenvalue, and stability penalties lets users specify fine distinctions such as stable versus unstable sinks and sources, or node versus focus, not just coarse point type.
  • Multiple critical points can be specified at once, and the achievable spacing between them is bounded by the modeled distribution, which the paper uses to estimate typical critical-point distances in the fluid ensemble.

Reading between the lines

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

  • Inference: the same energy-on-derivatives recipe should transfer to 3D vector fields and to scalar fields, since critical points and Morse-Smale cells are defined through the same Jacobian or Hessian machinery; the paper does not test this.
  • Inference: the alignment drop for two nearby critical points in Table 4 suggests the fluid-flow distribution has an effective minimum separation between same-sign critical points; comparing that spacing with vortex-core statistics in the simulation data would be a direct check.
  • Inference: because guidance is applied only after a chosen start time and uses the predicted clean latent, the method implicitly trades diversity for control; measuring Fréchet distance across guidance start times would quantify that tradeoff.
  • Inference: any smooth function of pointwise field derivatives could replace the topology energy and yield a new training-free control, so the approach generalizes to features like vorticity or other vortex criteria that the paper only mentions as future directions.
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

4 major / 5 minor

Summary. The paper introduces "topology guidance," a training-free method for steering the DDIM sampling process of a latent diffusion model over Functa/SIREN-coded 2D vector fields. Guidance energies are defined to enforce the presence of a critical point at a user-specified location, to control its type (sink, source, saddle) through eigenvalue-sign penalties, and to control its stability through an additional eigenvalue-based term. The method is evaluated on a 2D fluid-flow ensemble, reporting alignment rates above 85% in most configurations while keeping Fréchet Distance close to the unguided baseline, together with qualitative studies of topology-conditioned distributions and a "fill-in-the-gap" analysis.

Significance. If the method works as claimed, it is a useful contribution to visualization and generative-model control: it enables users to specify topological features of generated vector fields without retraining, and the ensemble-comparison application is compelling. The paper has clear strengths: large-scale quantitative evaluation (2,500 samples per configuration in Table 1), distribution-fidelity checks via Fréchet Distance, multi-critical-point experiments, and an interesting qualitative analysis of how the diffusion model fills in the gap when the prescribed topology is moved. However, the central derivation currently contains load-bearing inconsistencies in the treatment of foci and stability, so the reported results cannot be attributed to the algorithm as written.

major comments (4)
  1. [Sec. 4.3, Eq. (12)] The statement that the sign of Δ in Eq. (12) determines stability is incorrect. Δ = Tr(J)^2 − 4|J| distinguishes nodes (real eigenvalues, Δ > 0) from foci (complex conjugate eigenvalues, Δ < 0); stability is determined by the real parts of the eigenvalues, which in 2D is governed by Tr(J). As written, the stability energy in Eq. (13) cannot separate stable from unstable foci.
  2. [Sec. 4.3, Eq. (13)] Minimizing E_s = σ(β_s λ_1) gives the opposite of the stated labels: for "stable" with β_s = −1, σ(−λ_1) is minimized as λ_1 → +∞, driving the eigenvalue positive; for "unstable" with β_s = 1, minimization drives λ_1 → −∞. Moreover, in the combined energy Eq. (14), the stability term σ(−λ_1) for a stable point directly conflicts with the type term σ(λ_1) for an attracting point. The sign convention in the text is reversed with respect to the stated minimization objective.
  3. [Sec. 4.2, Eqs. (10)-(11), and Fig. 4] Foci cannot be encoded by Eq. (10), because their eigenvalues are complex conjugates and do not admit the real ordering used there; substituting complex λ into the sigmoid in Eq. (11) does not yield a real energy. Figure 4 lists σ(Δ) terms for focus and node configurations, but Eq. (14) omits any Δ term. The focus and node results in Tables 1 and 3 therefore are not produced by the energy functions stated in the paper. The authors must state the actual implemented energies (e.g., using real parts of eigenvalues or σ(Δ) terms) and confirm that the reported evaluation uses those corrected energies.
  4. [Sec. 5.1 and Algorithm 1] The values of the guidance strength ω and the guidance-start timestep are not specified; only Fig. 3 shows an injection at t = 600 for one example. These are free parameters that materially affect the alignment/fidelity trade-off reported in Tables 1-3, so without them Algorithm 1 is not reproducible.
minor comments (5)
  1. [Eq. (17)] The formula for Fréchet Distance appears to be missing an opening norm bar: it should read ∥μ1 − μ2∥^2, not "μ1− μ2∥2".
  2. [Sec. 4.3] The sentence "otherwise we have negative critical point" is a typo; it should say "unstable critical point".
  3. [Fig. 4] The text in Fig. 4 is garbled in several places, with missing square-root symbols and mismatched labels between configurations and energy terms; the figure should be redrawn so that each critical-point configuration and its associated constraints are unambiguous.
  4. [Sec. 5.2] The FD metric is evaluated on latent embeddings, but the embedding network is not described; please specify how the Fréchet Distance is computed on the latent distributions.
  5. [Sec. 4.1] The claim that the predicted clean latent ẑ_t is sufficiently in-distribution for the SIREN is plausible but not quantitatively validated; a small experiment reporting reconstruction error of ẑ_t versus t, or the norm of the guidance gradient, would strengthen this key assumption.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the guidance energies are explicit optimization objectives, and the reported alignment and FD metrics are empirical checks rather than re-statements of the method's inputs.

full rationale

The paper's derivation chain is self-contained in the relevant sense. The guidance functions in Eqs. (7)-(8), (11), (13), and (14) are explicitly written as differentiable objectives over the SIREN's output value and Jacobian eigenvalues, combined with the standard DDIM guidance update in Eq. (6). The claimed capability, namely that generated fields contain critical points of a prescribed location and type, is not an input to these equations; it is an outcome obtained by optimizing the latent trajectory through gradient descent. The quantitative evaluation checks this outcome using a separate grid-based zero-crossing extraction procedure (Sec. 5.1), and it reports alignment below 100% in several configurations, which confirms that the metric is not satisfied by construction. The FD metric in Eq. (17) compares generated latents to the real ensemble latent distribution, providing an external fidelity check rather than a tautological one. No parameter is fitted to the evaluation metric and then reported as a prediction; the guidance strength and guidance start time are hyperparameters determined by experimentation, not fitted predictions. The paper's citations to Functa, SIREN, DDPM, and diffusion guidance are standard external building blocks, and there is no load-bearing self-citation chain or invoked uniqueness theorem. The skeptical concerns about Eq. (13)'s sign convention and the treatment of complex eigenvalues for focus configurations are substantive correctness risks, but they are not circularity: they assert that the stated energy may not implement the intended stability control, not that the paper's result is equivalent to its input by construction. Overall, the central claims are empirically testable and are tested against independently defined criteria, so no circular step is present.

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

The method introduces no new physical entities or conserved quantities. It relies on trained neural components (SIREN and diffusion model) and a set of hand-designed guidance energy terms. The main free parameters are the guidance strength, the start timestep, and the beta signs in the eigenvalue sigmoids. The central assumptions are that the SIREN derivatives are faithful, that predicted clean latents stay in-distribution enough for guidance, and that the latent space is smooth enough for gradient-based steering.

free parameters (3)
  • guidance_strength_omega = not reported
    Hyperparameter omega in Eq. (6) controls the trade-off between topology adherence and data fidelity; no value or selection procedure is given in the text.
  • guidance_start_timestep = 600 (per Fig. 3)
    Chosen through experimentation based on when the norm begins to decrease; this hand-tuned parameter affects all results and is not derived from first principles.
  • eigenvalue_sigmoid_signs_beta = +1 or -1
    The beta values in Eqs. (11)-(13) are set to +/-1 to encode the target type and stability. They are design choices, not fitted to the data distribution, but they are free parameters of the guidance formulation.
assumptions (4)
  • domain assumption The Functa SIREN provides a sufficiently accurate and differentiable representation of the ensemble fields, including spatial derivatives needed for Jacobian-based topology classification.
    Invoked throughout Sec. 4; the method depends on the SIREN Jacobian being faithful to the true field Jacobian. Fig. 7 reports PSNR evidence for field values, but no direct validation of derivative fidelity is provided.
  • domain assumption The predicted clean latent z_t (Eq. 5) is sufficiently in-distribution for the SIREN to yield meaningful guidance gradients at the chosen timesteps.
    Stated as a problem in Sec. 4.1: 'When t is large, we find that the supplied vector zt is out-of-distribution for the SIREN...'. The method assumes the DDIM-based z_t fixes this sufficiently; no quantitative check is given.
  • standard math The 2x2 Jacobian eigenvalue classification applies to the SIREN-represented fields at the prescribed points, and the guidance energy can be minimized to reach the target class.
    Standard linearization of critical points in 2D vector fields, used in Eqs. (9)-(13). This is background math and is not at issue, but it is an unproved background result.
  • domain assumption The diffusion model latent space is smooth enough that gradients of the energy with respect to z_t produce local changes in the decoded field without destroying global structure.
    The entire guidance method relies on this smoothness; the paper provides only indirect evidence via qualitative results in Figs. 8-10.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Topology Guidance: Controlling the Outputs of Generative Models via Vector Field Topology." pith.science (2026). https://pith.science/paper/IN7EAK3N

@misc{pith2026250506804,
  author       = {Pith},
  title        = {Pith review of: Topology Guidance: Controlling the Outputs of Generative Models via Vector Field Topology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IN7EAK3N}},
  note         = {Machine review of arXiv:2505.06804}
}
read the original abstract

For domains that involve numerical simulation, it can be computationally expensive to run an ensemble of simulations spanning a parameter space of interest to a user. To this end, an attractive surrogate for simulation is the generative modeling of fields produced by an ensemble, allowing one to synthesize fields in a computationally cheap, yet accurate, manner. However, for the purposes of visual analysis, a limitation of generative models is their lack of control, as it is unclear what one should expect when sampling a field from a model. In this paper we study how to make generative models of fields more controllable, so that users can specify features of interest, in particular topological features, that they wish to see in the output. We propose topology guidance, a method for guiding the sampling process of a generative model, specifically a diffusion model, such that a topological description specified as input is satisfied in the generated output. Central to our method, we couple a coordinate-based neural network used to represent fields, with a diffusion model used for generation. We show how to use topologically-relevant signals provided by the coordinate-based network to help guide the denoising process of a diffusion model. This enables us to faithfully represent a user's specified topology, while ensuring that the output field remains within the generative data distribution. Specifically, we study 2D vector field topology, evaluating our method over an ensemble of fluid flows, where we show that generated vector fields faithfully adhere to the location, and type, of critical points over the spatial domain. We further show the benefits of our method in aiding the comparison of ensembles, allowing one to explore commonalities and differences in distributions along prescribed topological features.

Figures

Figures reproduced from arXiv: 2505.06804 by the authors.

Figure 1
Figure 1. We show topology-guided sampling, for different distributions formed over a simulation ensemble. (a) Generating a sink in the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. We present an overview of the model we use for topology guid [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Localized effect of guidance on vector field generation. Left: [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Our method provides a general framework for drawing vector fields [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Manipulating critical point types using an energy function inspired [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: An energy function that applied to norms, eigenvalues, and [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: PSNR distribution of training and test datasets for latent repre [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Example samplings when prescribing different topology configura [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 10
Figure 10. Figure 10: Exploring the diffusion model’s “fill in the gap" strategy by speci [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 9
Figure 9. Figure 9: Comparing topology structure distribution of different guidance [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 45 canonical work pages

  1. [1]

    Bansal, H.-M

    A. Bansal, H.-M. Chu, A. Schwarzschild, S. Sengupta, M. Goldblum, J. Geiping, and T. Goldstein. Universal guidance for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 843–852, 2023. 4, 5

  2. [2]

    Bauer, E

    M. Bauer, E. Dupont, A. Brock, D. Rosenbaum, J. R. Schwarz, and H. Kim. Spatial functa: Scaling functa to imagenet classification and generation, Feb. 2023. 8, 9

  3. [3]

    M. J. Buehler. Predicting mechanical fields near cracks using a progressive transformer diffusion model and exploration of generalization capacity. Journal of Materials Research, 38(5):1317–1331, 2023. 4

  4. [4]

    G. Chen, V . Kwatra, L.-Y . Wei, C. D. Hansen, and E. Zhang. Design of 2d time-varying vector fields. IEEE Transactions on Visualization and Computer Graphics, 18(10):1717–1730, 2011. 3

  5. [5]

    Cheng, A

    H.-C. Cheng, A. Cardone, S. Jain, E. Krokos, K. Narayan, S. Subramaniam, and A. Varshney. Deep-learning-assisted volume visualization. IEEE transactions on visualization and computer graphics, 25(2):1378–1391,

  6. [6]

    Dallmann, T

    U. Dallmann, T. Herberg, G. H., S. W.H., and Z. H.Q. Flow field diagnostics-topological flow changes and spatio-temporal flow structure. In 33rd Aerospace Sciences Meeting and Exhibit, p. 791, 1995. 3

  7. [7]

    L. Deng, Y . Wang, Y . Liu, F. Wang, S. Li, and J. Liu. A cnn-based vortex identification method. Journal of Visualization, 22:65–78, 2019. 3

  8. [8]

    Dhariwal and A

    P. Dhariwal and A. Nichol. Diffusion models beat gans on image synthesis. Advances in neural information processing systems, 34:8780–8794, 2021. 2, 4

Show all 55 references
  1. [9]

    Drygala, B

    C. Drygala, B. Winhart, F. di Mare, and H. Gottschalk. Generative model- ing of turbulence. Physics of Fluids, 34(3), 2022. 3

  2. [10]

    Dupont, H

    E. Dupont, H. Kim, S. Eslami, D. Rezende, and D. Rosenbaum. From data to functa: Your data point is a function and you can treat it like one. arXiv preprint arXiv:2201.12204, 2022. 2, 4, 7, 8, 9

  3. [11]

    Epstein, A

    D. Epstein, A. Jabri, B. Poole, A. Efros, and A. Holynski. Diffusion self-guidance for controllable image generation. Advances in Neural Information Processing Systems, 36, 2024. 4

  4. [12]

    C. Finn, P. Abbeel, and S. Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In International conference on machine learning, pp. 1126–1135. PMLR, 2017. 4

  5. [13]

    Günther and H

    T. Günther and H. Theisel. The state of the art in vortex extraction. In Computer Graphics Forum, vol. 37, pp. 149–173. Wiley Online Library,

  6. [14]

    Han and C

    J. Han and C. Wang. Surfnet: Learning surface representations via graph convolutional network. In Computer Graphics Forum, vol. 41, pp. 109–

  7. [15]

    Han and C

    J. Han and C. Wang. Coordnet: Data generation and visualization gen- eration for time-varying volumes via a coordinate-based neural network. IEEE Transactions on Visualization and Computer Graphics, 29(12):4951– 4963, Dec. 2023. doi: 10.1109/TVCG.2022.3197203 2

  8. [16]

    J. Han, H. Zheng, D. Z. Chen, and C. Wang. Stnet: An end-to-end generative framework for synthesizing spatiotemporal super-resolution volumes. IEEE Transactions on Visualization and Computer Graphics, 28(1):270–280, Jan. 2022. doi: 10.1109/TVCG.2021.3114815 2

  9. [17]

    M. S. Hassouna and A. A. Farag. On the extraction of curve skeletons using gradient vector flow. In 2007 IEEE 11th International Conference on Computer Vision, pp. 1–8. IEEE, 2007. 3

  10. [18]

    W. He, J. Wang, H. Guo, K.-C. Wang, H.-W. Shen, M. Raj, Y . S. G. Nashed, and T. Peterka. Insitunet: Deep image synthesis for parameter space exploration of ensemble simulations. IEEE Transactions on Visual- ization and Computer Graphics, pp. 1–1, 2019. doi: 10.1109/TVCG.2019 ...

  11. [19]

    J. L. Helman and L. Hesselink. Visualizing vector field topology in fluid flows. IEEE Computer Graphics and Applications, 11(3):36–46. doi: 10. 1109/38.79452 3, 6

  12. [20]

    J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models,

  13. [21]

    Ho and T

    J. Ho and T. Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022. 4

  14. [22]

    F. Hong, C. Lai, H. Guo, E. Shen, X. Yuan, and S. Li. Flda: Latent dirichlet allocation based unsteady flow analysis. IEEE transactions on visualization and computer graphics, 20(12):2545–2554, 2014. 3

  15. [23]

    T. Hsu, W. K. Epting, H. Kim, H. W. Abernathy, G. A. Hackett, A. D. Rollett, P. A. Salvador, and E. A. Holm. Microstructure generation via generative adversarial network for heterogeneous, topologically complex 3d materials. Jom, 73:90–102, 2021. 3

  16. [24]

    Jakob, M

    J. Jakob, M. Gross, and T. Günther. A fluid flow data set for machine learning and its application to neural flow map interpolation. IEEE Trans- actions on Visualization and Computer Graphics, 27(2):1279–1289, 2020. 2, 3, 7, 9

  17. [25]

    B. Jing, E. Erives, P. Pao-Huang, G. Corso, B. Berger, and T. S. Jaakkola. Eigenfold: Generative protein structure prediction with diffusion models. In ICLR 2023-Machine Learning for Drug Discovery workshop, 2023. 4

  18. [26]

    Kingma, T

    D. Kingma, T. Salimans, B. Poole, and J. Ho. Variational diffusion models. Advances in neural information processing systems , 34:21696–21707,

  19. [27]

    Kohl, L.-W

    G. Kohl, L.-W. Chen, and N. Thuerey. Benchmarking autoregressive conditional diffusion models for turbulent flow simulation, 2024. 1

  20. [28]

    R. S. Laramee, H. Hauser, L. Zhao, and F. H. Post. Topology-based flow visualization, the state of the art. Topology-based methods in visualization, pp. 1–19, 2007. 3

  21. [29]

    Laubscher and P

    R. Laubscher and P. Rousseau. Application of generative deep learning to predict temperature, flow and species distributions using simulation data of a methane combustor. International Journal of Heat and Mass Transfer, 163:120417, 2020. 3

  22. [30]

    Lee and G

    K.-H. Lee and G. J. Yun. Microstructure reconstruction using diffusion- based generative models. Mechanics of Advanced Materials and Struc- tures, pp. 1–19, 2023. 3

  23. [31]

    S. Li, N. Marsaglia, C. Garth, J. Woodring, J. Clyne, and H. Childs. Data reduction techniques for simulation, visualization and data analysis. In Computer graphics forum, vol. 37, pp. 422–447. Wiley Online Library,

  24. [32]

    Lienen, D

    M. Lienen, D. Lüdke, J. Hansen-Palmus, and S. Günnemann. From Zero to Turbulence: Generative Modeling for 3D Flow Simulation. In International Conference on Learning Representations, 2024. 3, 4

  25. [33]

    Pobitzer, R

    A. Pobitzer, R. Peikert, R. Fuchs, B. Schindler, A. Kuhn, H. Theisel, K. Matkovi ´c, and H. Hauser. The state of the art in topology-based visualization of unsteady flow. In Computer Graphics Forum, vol. 30, pp. 1789–1811. Wiley Online Library, 2011. 3

  26. [34]

    Rombach, A

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer. High- resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 10684–10695, 2022. 4

  27. [35]

    Santurkar, D

    S. Santurkar, D. Budden, and N. Shavit. Generative compression. In 2018 Picture Coding Symposium (PCS), pp. 258–262. IEEE, 2018. 3

  28. [36]

    Shen and H.-W

    J. Shen and H.-W. Shen. Psrflow: Probabilistic super resolution with flow-based models for scientific data. IEEE Transactions on Visualization and Computer Graphics, 30(1):986–996, Jan. 2024. doi: 10.1109/TVCG. 2023.3327171 2

  29. [37]

    N. Shi, J. Xu, H. Li, H. Guo, J. Woodring, and H.-W. Shen. Vdl-surrogate: A view-dependent latent-based model for parameter space exploration of ensemble simulations. IEEE Transactions on Visualization and Computer Graphics, 29(1):820–830, 2022. 2

  30. [38]

    N. Shi, J. Xu, S. W. Wurster, H. Guo, J. Woodring, L. P. Van Roekel, and H.- W. Shen. Gnn-surrogate: A hierarchical and adaptive graph neural network for parameter space exploration of unstructured-mesh ocean simulations. IEEE Transactions on Visualization and Computer Graphic...

  31. [39]

    J. Shin, V . Xing, M. Pfitzner, and C. Lapeyre. Probabilistic deep learning of turbulent premixed combustion. AIP Advances, 13(8), 2023. 3

  32. [40]

    Sitzmann, J

    V . Sitzmann, J. Martel, A. Bergman, D. Lindell, and G. Wetzstein. Implicit neural representations with periodic activation functions. In Advances in Neural Information Processing Systems, vol. 33, pp. 7462–7473. Curran Associates, Inc., 2020. 2, 4

  33. [41]

    J. Song, C. Meng, and S. Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020. 4

  34. [42]

    Takahashi, I

    S. Takahashi, I. Fujishiro, Y . Takeshima, and T. Nishita. A feature-driven approach to locating optimal viewpoints for volume visualization. In VIS

  35. [43]

    Tancik, P

    M. Tancik, P. P. Srinivasan, B. Mildenhall, S. Fridovich-Keil, N. Raghavan, U. Singhal, R. Ramamoorthi, J. T. Barron, and R. Ng. Fourier features let networks learn high frequency functions in low dimensional domains, June 2020. 2

  36. [44]

    IEEE Visualization, 2005., pp. 495–502. IEEE, 2005. 2

  37. [45]

    H. Theisel. Designing 2d vector fields of arbitrary topology. In Computer Graphics Forum, vol. 21, pp. 595–604. Wiley Online Library, 2002. 3

  38. [46]

    Tang and C

    K. Tang and C. Wang. Stsr-inr: Spatiotemporal super-resolution for mul- tivariate time-varying volumetric data via implicit neural representation. Computers & Graphics, 119:103874, Apr. 2024. doi: 10.1016/j.cag.2024. 01.001 2

  39. [47]

    P. Vincent. A connection between score matching and denoising autoen- coders. Neural computation, 23(7):1661–1674, 2011. 4

  40. [48]

    Theisel, T

    H. Theisel, T. Weinkauf, H.-C. Hege, and H.-P. Seidel. Topological methods for 2d time-dependent vector fields based on stream lines and path lines. IEEE Transactions on Visualization and Computer Graphics, 11(4):383–394, 2005. 3

  41. [49]

    J. Wang, S. Hazarika, C. Li, and H.-W. Shen. Visualization and visual analysis of ensemble data: A survey. IEEE transactions on visualization and computer graphics, 25(9):2853–2872, 2018. 2

  42. [50]

    Wang and J

    C. Wang and J. Han. Dl4scivis: A state-of-the-art survey on deep learn- ing for scientific visualization. IEEE transactions on visualization and computer graphics, 2022. 2

  43. [51]

    S. W. Wurster, H. Guo, H.-W. Shen, T. Peterka, and J. Xu. Deep hierarchi- cal super resolution for scientific data. IEEE Transactions on Visualization and Computer Graphics, 2022. 2

  44. [52]

    Weinkauf, H

    T. Weinkauf, H. Theisel, K. Shi, H.-C. Hege, and H.-P. Seidel. Extracting higher order critical points and topological simplification of 3d vector fields. In VIS 05. IEEE Visualization, 2005., pp. 559–566. IEEE, 2005. 3

  45. [53]

    A. Zhou, K. Yang, Y . Jiang, K. Burns, W. Xu, S. Sokota, J. Z. Kolter, and C. Finn. Neural functional transformers. Advances in Neural Information Processing Systems, 36, 2024. 9

  46. [54]

    Zhang, J

    E. Zhang, J. Hays, and G. Turk. Interactive tensor field design and visual- ization on surfaces. IEEE Transactions on Visualization and Computer Graphics, 13(1):94—-107, jan 2007. doi: 10.1109/TVCG.2007.16 3

  47. [120]

    Wiley Online Library, 2022. 2

Pith tools

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