Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Geometry Matters: Benchmarking Scientific ML Approaches for Flow Prediction around Complex Geometries

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read On a 3,000-simulation complex-geometry flow benchmark, transformer-based operator models scOT and Poseidon beat classical neural operators, with the largest gap in low-data settings.

desk verdict The data-efficiency benchmark is real and useful, but the geometry-generalization claim is unsupported because the random split shares geometries between train and test. read the letter →

arxiv 2501.01453 v2 pith:S2K6EAYX submitted 2024-12-31 cs.LG physics.flu-dyn

Ali Rabeh , Ethan Herron , Aditya Balu , Soumik Sarkar , Chinmay Hegde , Adarsh Krishnamurthy , Baskar Ganapathysubramanian This is my paper · ORCID
classification cs.LGphysics.flu-dyn
keywords ScientificmachinelearningNeuraloperatorsVisiontransformersSigneddistancefieldsBinarymasksLid-drivencavityflowOut-of-distributiongeneralizationDataefficiency
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 sets out to show that for predicting steady fluid flow around intricate two-dimensional geometries, the model class and the geometry encoding should be chosen together. Using the FlowBench 2D lid-driven cavity subset—3,000 simulations over 300 complex shapes—it benchmarks five classical neural operators against six variants of the transformer-based scOT/Poseidon family, under two geometry representations (signed distance fields and binary masks), several training-set sizes, and a Reynolds-number extrapolation split. Its central empirical claim is that scOT and Poseidon are the most accurate models, often by an order of magnitude in mean squared error, and that their advantage is most pronounced when training data is scarce, with pretrained Poseidon-T reaching roughly $10^{-4}$ MSE below 800 samples. It also claims that neither geometry representation is universally best: binary masks help scOT, Poseidon, and CNO, while SDF helps the other neural operators. If correct, the results give practitioners a concrete selection rule—transformer-based operators for data-limited complex-geometry flow prediction, with representation tuned per architecture—and a unified scoring scale for comparing future models.

What carries the argument

The argument is carried by a benchmark design and a scoring device. The design pairs each of 300 complex geometries with 10 Reynolds numbers, feeds the model a Reynolds scalar plus either a binary mask or a signed distance field at $512 \times 512$ resolution, and asks it to output velocity components and pressure. The models are grouped into neural operators—FNO, CNO, WNO, DeepONet, and geometric-DeepONet—and six scOT/Poseidon variants; scOT is a hierarchical multiscale vision transformer built from Swin Transformer V2 and ConvNeXt residual blocks in a U-Net encoder-decoder, and Poseidon is the same architecture pretrained on PDE datasets. The scoring device is a logarithmic score mapping MSE $\text{max}=1$ (a zero-prediction baseline) to 0 and MSE $\text{min}=10^{-6}$ (CFD numerical accuracy) to 100, applied to global MSE (M1), boundary-layer MSE inside SDF $\le 0.2$ (M2), and the $\ell^2$ norm of Navier-Stokes momentum residuals (M3). This unified scale is what lets the paper compare models across accuracy, near-wall behavior, and physical consistency.

What would settle it

Retrain scOT-T, Poseidon-T, CNO, and DeepONet on a geometry-stratified split that holds out entire geometries or entire shape families instead of individual simulations, and compare held-out-shape MSE; if the transformer-based margin disappears or reverses, the central generalization claim is refuted.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is a measured performance ranking with an architecture-dependent geometry-encoding result. Across random and extrapolatory test splits, scOT and Poseidon consistently deliver the lowest global and boundary-layer error, with Poseidon-T and scOT-T reaching scores around 65 on the random split (MSE near $10^{-4}$) while FNO, CNO, DeepONet, and WNO lag well behind; pretraining gives Poseidon its largest advantage when only a few hundred training samples are available, and by full data scOT matches it. The paper further finds that binary masks improve scOT, Poseidon, and CNO, whereas signed distance fields improve the other neural operators, and that neural operators saturate near 800 training samples with the mask representation while transformer-based models continue to improve with more data. On the physical-consistency metric, DeepONet records the lowest momentum residual despite worse pointwise accuracy. Finally, all models degrade sharply on the extrapolatory split, where test Reynolds numbers lie in the top or bottom 10 percent of the training range, showing that out-of-distribution generalization remains unsolved.

Load-bearing premise

The load-bearing assumption is that the random 80/20 split of the 3,000 individual flow samples measures a model's ability to handle complex geometries, even though each of the 300 shapes contributes 10 Reynolds-number simulations, so the same shapes appear in both training and test sets; the extrapolatory split tests new Reynolds numbers, not new geometry.

Editorial extensions

If this is right

  • A practitioner with a small CFD dataset around complex 2D shapes can expect scOT or Poseidon to reach usable accuracy around $10^{-4}$ MSE with a few hundred samples, where classical neural operators are roughly an order of magnitude worse.
  • Geometry representation should be chosen per architecture: binary masks for scOT, Poseidon, and CNO; signed distance fields for FNO, DeepONet, geometric-DeepONet, and WNO.
  • Neural operators trained with binary masks stop improving near 800 samples, so extra data for them has little payoff; transformer-based models can still convert additional samples into lower error.
  • DeepONet remains the best choice when satisfying the momentum equation matters more than pointwise accuracy, and it is also the fastest at inference.
  • Out-of-distribution Reynolds numbers defeat all models regardless of training-set size, so deployment beyond the training range needs a different strategy.

Reading between the lines

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

  • An editorial check: the paper's random 80/20 split is over individual simulations, and since each of the 300 geometries contributes 10 Reynolds numbers, the same shapes appear in both training and test sets; this means the reported accuracy may overstate generalization to unseen geometries, and a geometry-stratified holdout is needed to confirm the ranking.
  • If the data-efficiency result transfers beyond lid-driven cavities, pretrained operator transformers could serve as general-purpose initializations for steady-flow surrogates, cutting the cost of building new CFD surrogates.
  • The M3 residual metric appears to reward smoother fields, so combining it with gradient errors—as the paper's appendix does—may better capture the accuracy that lift and drag calculations require.
  • A direct testable extension would be to hold out entire shape families (NURBS, spherical harmonics, skeleton-derived shapes) and fine-tune Poseidon on two families while testing on the third; this isolates whether the advantage comes from geometric generalization or from memorizing seen shapes.
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 / 4 minor

Summary. The paper benchmarks 11 scientific-machine-learning models on the FlowBench 2D lid-driven cavity dataset (300 geometries, 10 Reynolds numbers each, 3000 samples). It compares SDF and binary-mask geometry encodings, studies training-set-size scaling, and evaluates random versus Reynolds-extreme 'extrapolatory' splits. A unified 0-100 score combines global MSE, boundary-layer MSE, and PDE momentum residual. The headline claims are that scOT/Poseidon foundation models outperform neural operators, especially in low-data regimes, and that SDF representations become superior with sufficient data, while all models struggle out of distribution.

Significance. The question is timely and the dataset/code release is a useful service. The paper is an honest attempt to compare many architectures under a common protocol, and the appendix provides substantial per-model detail. However, the current experimental design does not test geometry generalization because the sample-wise split leaks geometry identity; the SDF claim contradicts the main tables; and single-run comparisons lack uncertainty quantification. With a geometry-stratified split and revised claims, the benchmark could be a valuable reference for the SciML community.

major comments (4)
  1. [Section 2.1, Section 4.1] The 80/20 random split is sample-wise, not geometry-stratified. Since each of the 300 geometries contributes 10 Reynolds-number simulations, only about 32 geometries are completely held out (0.8^10 per geometry); the remaining ~268 geometries appear in both train and test. Consequently, the 'Random' results in Tables 1 and 2 measure interpolation/extrapolation in Reynolds number for mostly seen geometries, not generalization to unseen complex geometries. The 'extrapolatory' split in Section 2.3 also varies only Reynolds number, so it does not test geometry extrapolation either. The geometry-generalization claims in the abstract and Section 3 are therefore not supported by the reported experiments. Please add a split that holds out all 10 samples for a disjoint set of geometries and report those scores separately.
  2. [Abstract, Section 2.1, Tables 1 and 2] The claim that 'SDF representations yield superior results with sufficient training data' is contradicted by the full-data random scores: poseidon-T scores 66.0 M1 with mask versus 64.9 with SDF, scOT-T 67.8 versus 64.6, and CNO 47.3 versus 44.8; SDF is better only for some neural operators (e.g., FNO, geometric-DeepONet). The conclusion that the best geometric representation is model-dependent is supported, but the global SDF-superiority claim should either be removed or replaced by a per-model, per-metric analysis.
  3. [Sections 2.2 and 2.4, Tables 1-2 and A.1-A.8] All results appear to be single runs, with no random seeds, repetitions, or confidence intervals described in the experimental protocol. Many of the headline comparisons are small (e.g., scOT-T versus poseidon-T differ by 0.3-1.8 points in M1), so the statement that scOT/Poseidon 'consistently deliver the best performance, often by a significant margin' is not statistically supported. Please provide multiple seeds or error bars, or at least state explicitly that the margins are not significance-tested.
  4. [Section 2.2, Figure 4] The data-sufficiency experiments reduce the number of samples by drawing random subsets of the already-leaked training split, so the smaller configurations simultaneously reduce the number of geometries and the Reynolds coverage per geometry. The claim that 'neural operators reach an asymptotic error limit around 800 samples' conflates sample count with geometry diversity. The authors should report the number of distinct geometries and the Reynolds range covered at each training size, or stratify the training subsets by geometry.
minor comments (4)
  1. [Appendix A] Appendix A says the results are for 'one-third and one-tenth of the original data', but Table A.1 is labeled 'a subset of half of the dataset'; reconcile the labels and fix the 'exrapolatory' typo.
  2. [Section 4.3] The score's endpoints MSEmax=1 and MSEmin=1e-6 are hand-chosen; because the score is a monotone transform of MSE, rank order per metric is unaffected, but absolute statements like 'score=50' should be presented as conventions rather than calibrated accuracy.
  3. [Tables 1, 2, A.1-A.8] Several tables contain repeated 0.0 entries for WNO's M3 metric; please clarify whether these are genuine residual values or placeholder scores from predictions that could not be evaluated.
  4. [Figure A.2 caption] The caption says 'sample size of 300' in the third row while the text and tables refer to 240 samples; correct the inconsistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the benchmark conclusions are empirical and do not reduce to their inputs.

full rationale

The paper's central claims are that scOT and Poseidon outperform neural operators and that representation choice is model-dependent. These are empirical benchmark results obtained by training 11 models on the FlowBench 2D lid-driven cavity dataset and evaluating them with explicit metrics. No target quantity is fitted and then relabeled as a prediction: the scoring formula is a monotone logarithmic rescaling of MSE with hand-specified anchors (MSEmax=1, MSEmin=1e-6), so it does not by construction force any model ranking. The paper cites FlowBench, which is a self-citation from the same group, but FlowBench is a publicly available, externally generated CFD dataset; using it as a benchmark is not circular. The extrapolation experiments split only by Reynolds number, so the geometry-generalization interpretation is a validity concern, not a circularity: the reported numbers do not become equal to their inputs by definition. No uniqueness theorem, ansatz, or fitted parameter is imported from self-citations in a load-bearing way. The inconsistency between the abstract's blanket SDF claim and Tables 1-2 is a correctness issue, not a circular derivation. The paper is therefore self-contained for the purpose of circularity analysis.

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

The central benchmark rests on dataset ground truth, metric normalization choices, and a sample-splitting protocol. No physical constants are fitted and no new physical entities are introduced; the main structural risk is that the split does not isolate unseen-geometry generalization.

free parameters (2)
  • Score normalization bounds = MSEmax=1, MSEmin=1e-6
    Section 4.3; chosen by hand to map MSE to a 0-100 scale. They set the scale of reported scores but not the ranking.
  • Boundary layer threshold = SDF in [0, 0.2]
    Section 4.3; chosen by hand to define the M2 near-surface region. It affects M2 values only.
assumptions (3)
  • ad hoc to paper The 80/20 sample-wise split represents independent flow scenarios and tests geometry generalization.
    Section 2.1 uses a random sample split; Section 4.1 reports 10 simulations per geometry, so the split likely shares geometries between train and test.
  • domain assumption FlowBench CFD simulations are accurate enough to serve as ground truth.
    The paper uses the FlowBench dataset [19] as a high-fidelity reference and does not validate it against another solver in this work.
  • domain assumption Steady-state Navier-Stokes is the correct governing model for the lid-driven cavity benchmark.
    Section 4.1 defines the problem; the M3 residual is computed from the momentum equations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Geometry Matters: Benchmarking Scientific ML Approaches for Flow Prediction around Complex Geometries." pith.science (2026). https://pith.science/paper/S2K6EAYX

@misc{pith2026250101453,
  author       = {Pith},
  title        = {Pith review of: Geometry Matters: Benchmarking Scientific ML Approaches for Flow Prediction around Complex Geometries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S2K6EAYX}},
  note         = {Machine review of arXiv:2501.01453}
}
read the original abstract

Rapid and accurate simulations of fluid dynamics around complicated geometric bodies are critical in a variety of engineering and scientific applications, including aerodynamics and biomedical flows. However, while scientific machine learning (SciML) has shown considerable promise, most studies in this field are limited to simple geometries, and complex, real-world scenarios are underexplored. This paper addresses this gap by benchmarking diverse SciML models, including neural operators and vision transformer-based foundation models, for fluid flow prediction over intricate geometries. Using a high-fidelity dataset of steady-state flows across various geometries, we evaluate the impact of geometric representations -- Signed Distance Fields (SDF) and binary masks -- on model accuracy, scalability, and generalization. Central to this effort is the introduction of a novel, unified scoring framework that integrates metrics for global accuracy, boundary layer fidelity, and physical consistency to enable a robust, comparative evaluation of model performance. Our findings demonstrate that newer foundation models significantly outperform neural operators, particularly in data-limited scenarios, and that SDF representations yield superior results with sufficient training data. Despite these promises, all models struggle with out-of-distribution generalization, highlighting a critical challenge for future SciML applications. By advancing both evaluation models and modeling capabilities, our work paves the way for robust and scalable ML solutions for fluid dynamics across complex geometries.

Figures

Figures reproduced from arXiv: 2501.01453 by the authors.

Figure 1
Figure 1. A data-driven evaluation framework for accelerating PDE solvers of fluid flow around complex geometries using scientific machine learning models. This figure illustrates the scientific ML framework that assesses neural operators and foundation models for fluid flow solvers. Flow simulations are performed for steady-state lid-driven cavity flow across various complex geometries and Reynolds numbers. The top left subp… view at source ↗
Figure 2
Figure 2. Comparison of geometry representations for different geometries: (a-c) Signed Distance Field (SDF) rep￾resentations and (d-f) binary mask representations. SDF. Additionally, Tables 1 and 2 shows that scOT and Poseidon models outperform the other neural operators by roughly an order of magnitude. Sample field predictions and error comparisons between SDF and mask representa￾tions for the velocity in the y-direction o… view at source ↗
Figure 3
Figure 3. Comparison of score values for different models using Signed Distance Field (SDF) and binary mask representations. The bar plot shows the score for each model, indicating the performance difference between SDF and mask representations [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Comparison of score values vs. sample size for different scientific machine learning models using SDF and binary mask representations. The top row compares neural operators (CNO, FNO, Geo-DeepONet), while the bottom row compares foundation models (scOT-B, scOT-T, posei…
Figure 5
Figure 5. Figure 5: Histogram of Reynolds numbers for Train and Test splits in random and extrapolatory cases. 8 [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Comparison of score values for different models using random and extrapolatory test/train splits. The bar plot shows the score for each model, indicating the difference in performance between random and extrapolatory data splits. Poseidon-T, scOT-T, and CNO use a binar…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Defining Foundation Models for Computational Science: A Call for Clarity and Rigor

    cs.LG 2025-05 conditional novelty 4.0 of 10

    The paper defines foundation models for computational science and presents DD-FEM, a local-to-global data-driven framework inspired by finite elements, as a candidate path to meet that definition.

Reference graph

Works this paper leans on

51 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    Computational fluid dynamics of whole-body aircraft

    Ramesh Agarwal. Computational fluid dynamics of whole-body aircraft. Annual review of fluid mechanics, 31 (1):125–169, 1999

    work page 1999

  2. [2]

    Immersed boundary methods

    G Iaccarino and R Mittal. Immersed boundary methods. Annu. Rev. Fluid Mech, 37:239–261, 2005

    work page 2005

  3. [3]

    Review of research on vehicles aerodynamic drag reduction methods

    Mohd Nizam Sudin, Mohd Azman Abdullah, Shamsul Anuar Shamsuddin, Faiz Redza Ramli, and Musthafah Mohd Tahir. Review of research on vehicles aerodynamic drag reduction methods. International Journal of Mechanical and Mechatronics Engineering, 14(02):37–47, 2014

    work page 2014

  4. [4]

    A computational approach to modeling cellular-scale blood flow in complex geometry

    Peter Balogh and Prosenjit Bagchi. A computational approach to modeling cellular-scale blood flow in complex geometry. Journal of computational physics, 334:280–307, 2017

    work page 2017

  5. [5]

    Patient-specific modeling of geometry and blood flow in large arteries

    Luca Antiga. Patient-specific modeling of geometry and blood flow in large arteries. Politecnico di Milano, 2002

    work page 2002

  6. [6]

    Other geometries in architecture: bubbles, knots and minimal surfaces

    Tobias Wallisser. Other geometries in architecture: bubbles, knots and minimal surfaces. In Mathknow: Mathe- matics, Applied Sciences and Real Life, pages 91–111. Springer, 2009

    work page 2009

  7. [7]

    Deep learning for reduced order modelling and efficient temporal evolution of fluid simulations

    Pranshu Pant, Ruchit Doshi, Pranav Bahl, and Amir Barati Farimani. Deep learning for reduced order modelling and efficient temporal evolution of fluid simulations. Physics of Fluids, 33(10), 2021

    work page 2021

  8. [8]

    Recent advances and applications of deep learning methods in materials science

    Kamal Choudhary, Brian DeCost, Chi Chen, Anubhav Jain, Francesca Tavazza, Ryan Cohn, Cheol Woo Park, Alok Choudhary, Ankit Agrawal, Simon JL Billinge, et al. Recent advances and applications of deep learning methods in materials science. npj Computational Materials, 8(1):59, 2022

    work page 2022

Show all 51 references
  1. [9]

    Robust data-driven turbulence modeling for rans closures using a sciml approach for validation

    Uma Balakrishnan, William Rider, Matthew Barone, and Eric Parish. Robust data-driven turbulence modeling for rans closures using a sciml approach for validation. Bulletin of the American Physical Society, 2024

    2024

  2. [10]

    Mycrunchgpt: A llm assisted framework for scientific machine learning

    Varun Kumar, Leonard Gleyzer, Adar Kahana, Khemraj Shukla, and George Em Karniadakis. Mycrunchgpt: A llm assisted framework for scientific machine learning. Journal of Machine Learning for Modeling and Comput- ing, 4(4), 2023

    2023

  3. [11]

    The emergence and impact of scientific machine learning in geophysical exploration

    Umair bin Waheed. The emergence and impact of scientific machine learning in geophysical exploration. InThird International Meeting for Applied Geoscience & Energy, pages 1807–1812. Society of Exploration Geophysicists and American Association of Petroleum . . . , 2023

    2023

  4. [12]

    Scientific Machine Learning for Modeling and Discovery of Physical Systems with Quantified Uncertainty

    Luning Sun. Scientific Machine Learning for Modeling and Discovery of Physical Systems with Quantified Uncertainty. University of Notre Dame, 2023

    2023

  5. [13]

    Integrating scientific machine learning and physics-based models for quantification of uncertainty in thermal properties of silica aerogel

    Shayan Bhattacharjee. Integrating scientific machine learning and physics-based models for quantification of uncertainty in thermal properties of silica aerogel. Master’s thesis, State University of New York at Buffalo, 2023

    2023

  6. [14]

    Uncertainty quantifica- tion in scientific machine learning: Methods, metrics, and comparisons

    Apostolos F Psaros, Xuhui Meng, Zongren Zou, Ling Guo, and George Em Karniadakis. Uncertainty quantifica- tion in scientific machine learning: Methods, metrics, and comparisons. Journal of Computational Physics, 477: 111902, 2023

    2023

  7. [15]

    Florent Bonnet, Jocelyn Mazari, Paola Cinnella, and Patrick Gallinari. Airfrans: High fidelity computational fluid dynamics dataset for approximating reynolds-averaged navier–stokes solutions.Advances in Neural Information Processing Systems, 35:23463–23478, 2022

    2022

  8. [16]

    CFDBench: A large-scale benchmark for machine learning methods in fluid dynamics, 2024

    Yining Luo, Yingfa Chen, and Zhen Zhang. CFDBench: A large-scale benchmark for machine learning methods in fluid dynamics, 2024. arXiv preprint

    2024

  9. [17]

    MegaFlow2D: A parametric dataset for machine learning super-resolution in computational fluid dynamics simulations

    Wenzhuo Xu, Noelia Grande Gutierrez, and Christopher McComb. MegaFlow2D: A parametric dataset for machine learning super-resolution in computational fluid dynamics simulations. In Proceedings of Cyber- Physical Systems and Internet of Things Week 2023 . Association for Computi...

    2023

  10. [18]

    Rapid prediction of two-dimensional airflow in an operating room using scientific machine learning

    Gary Lynn Collins, Alexander New, Ryan A Darragh, Brian E Damit, and Christopher D Stiles. Rapid prediction of two-dimensional airflow in an operating room using scientific machine learning. In NeurIPS 2023 AI for Science Workshop, 2023

    2023

  11. [19]

    Flowbench: A large scale bench- mark for flow simulation over complex geometries

    Ronak Tali, Ali Rabeh, Cheng-Hau Yang, Mehdi Shadkhah, Samundra Karki, Abhisek Upadhyaya, Suriya Dhak- shinamoorthy, Marjan Saadati, Soumik Sarkar, Adarsh Krishnamurthy, et al. Flowbench: A large scale bench- mark for flow simulation over complex geometries. arXiv preprint arX...

    2024 arXiv

  12. [21]

    Effective geometric algorithms for im- mersed boundary method using signed distance field

    Chenguang Zhang, Chunliang Wu, and Krishnaswamy Nandakumar. Effective geometric algorithms for im- mersed boundary method using signed distance field. Journal of Fluids Engineering, 141(6):061401, 2019

    2019

  13. [22]

    Signed distance field enhanced fully resolved cfd-dem for simulation of granular flows involving multiphase fluids and irregularly shaped particles

    Zhengshou Lai, Jidong Zhao, Shiwei Zhao, and Linchong Huang. Signed distance field enhanced fully resolved cfd-dem for simulation of granular flows involving multiphase fluids and irregularly shaped particles. Computer Methods in Applied Mechanics and Engineering, 414:116195, 2023

    2023

  14. [23]

    Unit operation and process modeling with physics-informed machine learning

    Haochen Li, David Spelman, and John Sansalone. Unit operation and process modeling with physics-informed machine learning. Journal of Environmental Engineering, 150(4):04024002, 2024

    2024

  15. [24]

    Data-driven modeling of hypersonic reentry flow with heat and mass transfer

    Leonidas Gkimisis, Bruno Dias, James B Scoggins, Thierry Magin, Miguel A Mendez, and Alessandro Turchi. Data-driven modeling of hypersonic reentry flow with heat and mass transfer. AIAA Journal, 61(8):3269–3286, 2023

    2023

  16. [25]

    Computational challenges of viscous incompressible flows

    Dochan Kwak, Cetin Kiris, and Chang Sung Kim. Computational challenges of viscous incompressible flows. Computers & fluids, 34(3):283–299, 2005

    2005

  17. [26]

    Review of machine learning for hydrodynamics, transport, and reactions in multiphase flows and reactors

    Li-Tao Zhu, Xi-Zhong Chen, Bo Ouyang, Wei-Cheng Yan, He Lei, Zhe Chen, and Zheng-Hong Luo. Review of machine learning for hydrodynamics, transport, and reactions in multiphase flows and reactors. Industrial & Engineering Chemistry Research, 61(28):9901–9949, 2022

    2022

  18. [27]

    Towards foundation models for scientific machine learning: Characterizing scaling and transfer behavior

    Shashank Subramanian, Peter Harrington, Kurt Keutzer, Wahid Bhimji, Dmitriy Morozov, Michael W Mahoney, and Amir Gholami. Towards foundation models for scientific machine learning: Characterizing scaling and transfer behavior. Advances in Neural Information Processing Systems,...

    2024

  19. [28]

    Interpretable models for extrapolation in scientific machine learning

    Eric S Muckley, James E Saal, Bryce Meredig, Christopher S Roper, and John H Martin. Interpretable models for extrapolation in scientific machine learning. Digital Discovery, 2(5):1425–1435, 2023

    2023

  20. [29]

    Learning stiff chemical kinetics using extended deep neural operators

    Somdatta Goswami, Ameya D Jagtap, Hessam Babaee, Bryan T Susi, and George Em Karniadakis. Learning stiff chemical kinetics using extended deep neural operators. Computer Methods in Applied Mechanics and Engineering, 419:116674, 2024

    2024

  21. [30]

    Normalizing flows for probabilistic modeling and inference

    George Papamakarios, Eric Nalisnick, Danilo Jimenez Rezende, Shakir Mohamed, and Balaji Lakshmi- narayanan. Normalizing flows for probabilistic modeling and inference. Journal of Machine Learning Research, 22(57):1–64, 2021

    2021

  22. [31]

    A transdisciplinary review of deep learning research and its relevance for water resources scientists

    Chaopeng Shen. A transdisciplinary review of deep learning research and its relevance for water resources scientists. Water Resources Research, 54(11):8558–8593, 2018

    2018

  23. [32]

    Modeling and simulations of high-density two-phase flows using projection-based cahn-hilliard navier-stokes equations

    Ali Rabeh, Makrand A Khanwale, Jonghyun Lee, and Baskar Ganapathysubramanian. Modeling and simulations of high-density two-phase flows using projection-based cahn-hilliard navier-stokes equations. arXiv preprint arXiv:2406.17933, 2024

    2024 arXiv

  24. [33]

    Neural pde solvers for irregular domains

    Biswajit Khara, Ethan Herron, Aditya Balu, Dhruv Gamdha, Chih-Hsuan Yang, Kumar Saurabh, Anushrut Jignasu, Zhanhong Jiang, Soumik Sarkar, Chinmay Hegde, et al. Neural pde solvers for irregular domains. Computer-Aided Design, 172:103709, 2024. 15

    2024

  25. [34]

    Neufenet: Neural finite element solutions with theoretical bounds for parametric pdes

    Biswajit Khara, Aditya Balu, Ameya Joshi, Soumik Sarkar, Chinmay Hegde, Adarsh Krishnamurthy, and Baskar Ganapathysubramanian. Neufenet: Neural finite element solutions with theoretical bounds for parametric pdes. Engineering with Computers , 40(5):2761–2783, October 2024. doi...

    2024 doi

  26. [35]

    The shifted boundary method for embedded domain computations

    Alex Main and Guglielmo Scovazzi. The shifted boundary method for embedded domain computations. part I: Poisson and stokes problems. Journal of Computational Physics, 372:972–995, 2018

    2018

  27. [36]

    Optimal surrogate boundary selection and scalability studies for the shifted boundary method on octree meshes

    Cheng-Hau Yang, Kumar Saurabh, Guglielmo Scovazzi, Claudio Canuto, Adarsh Krishnamurthy, and Baskar Ganapathysubramanian. Optimal surrogate boundary selection and scalability studies for the shifted boundary method on octree meshes. Computer Methods in Applied Mechanics and En...

    2024

  28. [37]

    The NURBS book

    Les Piegl and Wayne Tiller. The NURBS book. Springer Science & Business Media, 2012

    2012

  29. [38]

    A simple method for particle shape generation with spher- ical harmonics

    Deheng Wei, Jianfeng Wang, and Budi Zhao. A simple method for particle shape generation with spher- ical harmonics. Powder Technology, 330:284–291, 2018. ISSN 0032-5910. doi: https://doi.org/10. 1016/j.powtec.2018.02.006. URL https://www.sciencedirect.com/science/article/pii/ ...

    2018

  30. [39]

    Skelneton 2019: Dataset and challenge on deep learning for geometric shape understanding

    Ilke Demir, Camilla Hahn, Kathryn Leonard, Geraldine Morin, Dana Rahbani, Athina Panotopoulou, Amelie Fondevilla, Elena Balashova, Bastien Durix, and Adam Kortylewski. Skelneton 2019: Dataset and challenge on deep learning for geometric shape understanding. In Proceedings of t...

    2019

  31. [40]

    Pyramid u-network for skeleton extraction from shape points

    Rowel Atienza. Pyramid u-network for skeleton extraction from shape points. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, June 2019

    2019

  32. [41]

    Internal flows

    JP Johnston. Internal flows. Turbulence, pages 109–169, 2005

    2005

  33. [42]

    Laminar, transitional, and turbulent flows in rotor-stator cavi- ties

    Brian Launder, S ´ebastien Poncet, and Eric Serre. Laminar, transitional, and turbulent flows in rotor-stator cavi- ties. Annual review of fluid mechanics, 42(1):229–248, 2010

    2010

  34. [43]

    Poseidon: Efficient foundation models for pdes, 2024

    Maximilian Herde, Bogdan Raoni ´c, Tobias Rohner, Roger K¨appeli, Roberto Molinaro, Emmanuel de B ´ezenac, and Siddhartha Mishra. Poseidon: Efficient foundation models for pdes, 2024

    2024

  35. [44]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. 2021 IEEE/CVF International Conference on Computer Vision (ICCV) , pages 9992–10002, 2021. URL https://api.seman...

    2021

  36. [45]

    Swin transformer v2: Scaling up capacity and resolution, 2022

    Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, and Baining Guo. Swin transformer v2: Scaling up capacity and resolution, 2022. URL https: //arxiv.org/abs/2111.09883

    2022 arXiv

  37. [46]

    A convnet for the 2020s, 2022

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, and Saining Xie. A convnet for the 2020s, 2022. URL https://arxiv.org/abs/2201.03545

    2022 arXiv

  38. [47]

    Fourier neural operator for parametric partial differential equations, 2021

    Zongyi Li, Nikola Kovachki, Kamyar Azizzadenesheli, Burigede Liu, Kaushik Bhattacharya, Andrew Stuart, and Anima Anandkumar. Fourier neural operator for parametric partial differential equations, 2021

    2021

  39. [48]

    Convolutional neural operators for robust and accurate learning of pdes, 2023

    Bogdan Raoni ´c, Roberto Molinaro, Tim De Ryck, Tobias Rohner, Francesca Bartolucci, Rima Alaifari, Sid- dhartha Mishra, and Emmanuel de B ´ezenac. Convolutional neural operators for robust and accurate learning of pdes, 2023

    2023

  40. [49]

    Wavelet neural operator for solving parametric partial differential equations in computational mechanics problems

    Tapas Tripura and Souvik Chakraborty. Wavelet neural operator for solving parametric partial differential equations in computational mechanics problems. Computer Methods in Applied Mechanics and Engineer- ing, 404:115783, 2023. ISSN 0045-7825. doi: https://doi.org/10.1016/j.cm...

    2023

  41. [50]

    Learning nonlinear oper- ators via deeponet based on the universal approximation theorem of operators

    Lu Lu, Pengzhan Jin, Guofei Pang, Zhongqiang Zhang, and George Em Karniadakis. Learning nonlinear oper- ators via deeponet based on the universal approximation theorem of operators. Nature Machine Intelligence, 3: 218––229, 2021

    2021

  42. [51]

    Universal approximation to nonlinear operators by neural networks with arbitrary activation functions and its application to dynamical systems

    Tianping Chen and Hong Chen. Universal approximation to nonlinear operators by neural networks with arbitrary activation functions and its application to dynamical systems. IEEE transactions on neural networks, 6(4):911– 917, 1995

    1995

  43. [52]

    Geom-deeponet: A point-cloud-based deep operator network for field predictions on 3d parameterized geometries

    Junyan He, Seid Koric, Diab Abueidda, Ali Najafi, and Iwona Jasiuk. Geom-deeponet: A point-cloud-based deep operator network for field predictions on 3d parameterized geometries. Computer Methods in Applied Mechanics and Engineering, 429:117130, September 2024. ISSN 0045-7825....

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.