Pith. sign in

REVIEW 4 major objections 5 minor 33 references

Improving hp-Variational Physics-Informed Neural Networks for Steady-State Convection-Dominated Problems

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that adding a network-learned SUPG stabilization term and an adaptive hard-constraint boundary indicator to variational physics-informed neural networks substantially improves accuracy on convection-dominated…

desk verdict Two sensible, incremental VPINN extensions, but the reported L2 errors are sampled on a grid that cannot see layers of width 1e-8 or 1e-4, so the claimed accuracy gains are not actually demonstrated. read the letter →

arxiv 2411.09329 v1 pith:WDDN5RHM submitted 2024-11-14 math.NA cs.CEcs.LGcs.NA

Thivin Anandh , Divij Ghose , Himanshu Jain , Pratham Sunkad , Sashikumaar Ganesan , Volker John This is my paper · ORCID
classification math.NAcs.CEcs.LGcs.NA MSC 65N3065N3568T07
keywords convection-diffusion-reactionproblemsconvection-dominatedhp-VPINNsFastVPINNsSUPGstabilizationhard-constrainedDirichletboundaryconditionsadaptiveindicatorfunctionphysics-informedneuralnetworks
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 make variational physics-informed neural networks (hp-VPINNs) accurate on convection-dominated convection-diffusion-reaction problems, where true solutions have thin boundary layers and standard loss functionals struggle. It proposes two extensions: a streamline-upwind Petrov–Galerkin (SUPG) stabilization term with a spatially varying stabilization parameter predicted by the network, and an adaptive indicator function whose boundary steepness exponents are learned during training. On benchmark problems with diffusion coefficient $\varepsilon=10^{-8}$, the learned-stabilization version lowers the $L^2$ error from $1.693\times10^{-4}$ to $1.037\times10^{-4}$ for the outflow-layer problem and from $1.544\times10^{-4}$ to $9.043\times10^{-5}$ for the parabolic-layer problem; the adaptive indicator improves this further to $5.887\times10^{-5}$ and $6.007\times10^{-5}$, respectively. These values lie well below the errors reported for PINN and hp-VPINN baselines in the literature. If the result holds, it provides a practical route to resolving boundary layers without resolving their width on a mesh, and it removes two hand-tuning burdens in physics-informed neural networks.

What carries the argument

The load-bearing object is the hard-constrained ansatz $u^{\text{hard}}_{\text{NN}}(x) = j(x) + h(x) u_{\text{NN}}(x;\theta)$, where $j$ extends the Dirichlet data, $h$ vanishes on $\partial\Omega$ and is positive inside, and boundary conditions are therefore satisfied exactly. On top of this, the SUPG term $L_{\text{SUPG}} = \int_\Omega \tau(x)(b\cdot\nabla u + cu - f)(b\cdot\nabla v)\,dx$ adds streamline diffusion; the stabilization parameter is not a tuned constant but a second network output, $\tau(x) = \tau_g \tanh(50x)\tanh(50y)\tanh(50(1-x))\tanh(50(1-y))\,\sigma(\tau_{\text{NN}}(x;\theta))$, which is positive, spatially varying, and zero at the boundary. For the indicator, the mechanism is an exponential family with per-boundary exponents $\kappa_i = 10^{\text{learnable exponent}}$, which lets the network adapt the slope of $h$ near each boundary to the steepness of the solution's layers. These two mechanisms together carry the reported accuracy improvement.

What would settle it

Re-run the two benchmark problems with all hyperparameters chosen before any solution information is used and with errors reported from a fixed run rather than the best epoch; if the learned-$\tau$ and adaptive-indicator errors no longer beat the plain variational loss, the central claim is falsified. Alternatively, apply the same architecture to a problem with a curved or internal layer and check whether the per-boundary exponential indicator family keeps the error below the plain-loss level.

Watch

Extended reading notes

Core claim

The central claim is that two modifications to the hp-VPINN loss functional turn a method that historically struggles on convection-dominated problems into one that resolves boundary layers at $\varepsilon=10^{-8}$ to roughly $10^{-4}$ accuracy on the studied benchmarks. First, replacing the standard variational loss $L^{\text{hard}}_{\text{var}}$ with $L^{\text{hard}}_{\text{var}} + L_{\text{SUPG}}$, where $L_{\text{SUPG}} = \int_\Omega \tau(x)(b\cdot\nabla u + cu - f)(b\cdot\nabla v)\,dx$, and having the network predict a positive, spatially varying $\tau$ that vanishes on the boundary, consistently outperforms both the plain loss and the same loss with a constant, hand-tuned $\tau$. Second, in the hard-constraint ansatz $u^{\text{hard}}_{\text{NN}} = j(x) + h(x)u_{\text{NN}}$, letting $h$ be of the form $(1-e^{-\kappa_1 x})(1-e^{-\kappa_1 y})(1-e^{-\kappa_2(1-x)})(1-e^{-\kappa_2(1-y)})$ with $\kappa_i = 10^{\alpha}, 10^{\beta}$, and learning $\alpha,\beta$ (and a third exponent for the parabolic-layer problem) alongside the solution, gives another accuracy gain and is stable across $\varepsilon = 10^{-4}, 10^{-6}, 10^{-8}$. The paper reports these combined choices as producing a large increase in accuracy.

Load-bearing premise

The improvements are measured with hyperparameters (constant $\tau$, $\tau_g$, and indicator exponents) selected by searching against the known exact solution and reporting the best epoch and best run, so if those hyperparameters had to be fixed without knowing the solution the reported gains could shrink or vanish; a second assumption is that the per-boundary exponential indicator family can represent the layer geometry of general convection-dominated solutions.

Editorial extensions

If this is right

  • On the two studied benchmarks, the combined method reaches $L^2$ errors below $1.1\times10^{-4}$ for the outflow-layer problem and below $6.1\times10^{-5}$ for the parabolic-layer problem at $\varepsilon=10^{-8}$, below the reported PINN and hp-VPINN baselines.
  • A network-predicted stabilization parameter removes the need for a grid search over $\tau$, and its spatial variation concentrates stabilization where the solution has large streamline derivatives.
  • Learning the indicator exponents provides a way to satisfy Dirichlet conditions exactly while adapting the near-boundary slope to layers, which the paper argues is essential for inlet and boundary layers.
  • The learned indicator parameters are nearly independent of $\varepsilon$ across the range $10^{-4}$ to $10^{-8}$, suggesting the mechanism is capturing layer geometry rather than a particular diffusion value.
  • Adding only an $L^2$ weight-regularization term did not improve accuracy, indicating that the paper's route to accuracy is stabilization and boundary representation, not weight decay.

Reading between the lines

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

  • A natural extension is to compare the spatially varying $\tau(x,y)$ learned by the network with a-posteriori optimized SUPG parameters; if the patterns match, it would support the view that the network discovers the same stabilization distribution as classical parameter optimization.
  • Because the learned indicator exponents varied only weakly with the Péclet number, one could pretrain the indicator branch on a cheap small-$\varepsilon$ run and reuse it for nearby diffusion coefficients; the paper does not test this transfer.
  • The exponential indicator family used here assigns one exponent per boundary segment, so curved or interior layers would likely require a richer parameterization; the reported accuracy gains may not automatically transfer to non-axis-aligned layers.
  • The paper reports best-epoch and, in the constant-parameter tables, best-run errors after tuning hyperparameters against the known exact solution; a stricter protocol that withholds the exact solution during tuning would clarify the size of the practical gain.
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 proposes two extensions to the FastVPINNs framework for steady-state convection-dominated convection-diffusion-reaction problems in two dimensions. First, it adds a SUPG-style stabilization term to the variational loss and introduces a neural-network head that predicts a spatially varying stabilization parameter tau(x,y), scaled by a tunable factor tau_g. Second, it proposes to learn the exponents in a fixed functional family of indicator functions used for hard-constrained Dirichlet boundary conditions. The authors test both ideas on two manufactured-solution problems with epsilon = 1e-8 (Pout with outflow layers and Ppara with exponential and parabolic layers) and report L2 errors that are lower than those obtained with the plain variational loss and lower than PINN/hp-VPINN results from Frerichs-Mihov et al. (2024a). The central claim is that these mechanisms produce a 'tremendous increase of accuracy' for layer-dominated solutions.

Significance. If the reported accuracy gains are genuine and reproducible, the two architectural ideas would be a useful contribution to the growing literature on physics-informed neural networks for singularly perturbed problems. The learned-tau idea, in particular, is a natural way to avoid hand-tuning a stabilization parameter, and the adaptive indicator function addresses an underappreciated source of error in hard-constrained PINNs. The paper is clearly written and the experiments are easy to follow. However, the current evaluation protocol does not establish the central claim: the error metric in Eq. (14) cannot see the layers that motivate the whole study, and the hyperparameters are selected in-sample on the exact solutions of the test problems. The significance of the contribution therefore depends on whether these methodological issues can be fixed in a revision.

major comments (4)
  1. [Eq. (14), Sections 4.2 and 5] The L2_err metric in Eq. (14) is evaluated on a uniform 100x100 grid with spacing about 0.01. For Pout the outflow layers have width O(epsilon)=1e-8, and for Ppara the parabolic layers have width O(sqrt(epsilon))=1e-4; both are far below the grid spacing, so no test point lies inside the layers. Consequently, the values reported in Tables 5-8 measure only interior accuracy, while the layers are the defining difficulty of convection-dominated problems and the stated motivation of the paper. A completely wrong layer of width delta and O(1) amplitude contributes at least O(sqrt(delta)) to the true L2 error, roughly 1e-4 for Pout and 1e-2 for Ppara, which is comparable to or much larger than the claimed errors (1.0e-4 to 6.0e-5). The claim of a 'tremendous increase of accuracy' for layer resolution is therefore not supported by the reported metric. Please compute the true L2 error with layer-resolving numerical quadrature (or on a graded mesh aligned with the layers) and additionally report errors restricted to the layer regions.
  2. [Section 4.3 and Section 5.1, Tables 3-6] The hyperparameters tau, lambda, and tau_g are selected by grid search using the known exact solution of the same two test problems, as described in Section 4.3 and in the text accompanying Tables 5-6. This is in-sample selection, so the reported improvements over the plain loss are partially fitted to the test cases. Moreover, Section 4.3 states that the smallest error obtained during training is reported ('after each epoch and not necessarily after the final epoch'), which adds an optimistic best-epoch bias. Please adopt a validation protocol that does not use the test solution for tuning (e.g., a fixed a priori choice or a separate validation split) and report the error at a fixed final epoch, together with means and standard deviations over multiple independent runs.
  3. [Section 5.2, Tables 7-8] The adaptive-indicator results report a single L2_err value per configuration with no number of runs, no standard deviation, and no statement about whether the value is the best run or the average over runs. Since Section 4.3 demonstrates that the minimum error varies noticeably across runs even with identical parameters, the single-point comparisons in Tables 7-8 are insufficient to establish that the learned alpha, beta, and gamma consistently improve accuracy. Please add multi-seed statistics and explicitly state the selection rule for the reported values.
  4. [Section 4.3 and Section 5, Tables 3-6] The comparison with the literature is limited to the PINN and hp-VPINN results from Frerichs-Mihov et al. (2024a). No comparison with a standard stabilized finite element method is provided, even though the MUAS method is used to generate the reference solution for Ppara and is mentioned as the currently most promising approach in the introduction. The two manufactured-solution problems share a similar structure (square domain, axis-aligned layers), so the generality of the conclusions is not yet demonstrated. Adding a stabilized FEM baseline and at least one problem with a different layer configuration would substantially strengthen the evidence for the paper's central claim.
minor comments (5)
  1. [Section 5.1, Tables 5-6] The column header 'best L2_err' conflicts with the text stating that the average L2_err values from 10 runs are presented; please clarify which value is actually reported in each table.
  2. [Section 4.1] For Ppara, the MUAS method is used to obtain a high-accuracy reference solution, but the accuracy of that reference (e.g., the mesh size and the resulting error) is never quantified; please state the grid and the estimated error of the reference solution.
  3. [Section 2 and Section 4.3] The extension function j(x) in the hard-constrained ansatz (6) is not given explicitly for Pout and Ppara; providing the formulas used for j(x) is necessary for reproducibility.
  4. [Figure 6] The text states that the plotted values are averages over 5 runs, but the figure does not show error bars or the spread of the data; adding them would make the sensitivity analysis more informative.
  5. [Section 5.1] The paper says that tau is set to zero on the boundary, but the indicator w(x) = tanh(50x)tanh(50y)tanh(50(1-x))tanh(50(1-y)) is only approximately zero there; please state whether quadrature nodes exactly on the boundary are excluded from the SUPG loss or whether the approximation is accepted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: both learned quantities are trained on PDE residuals, not on the reported L2-error metric.

full rationale

The paper's two proposals (learned stabilization parameter tau and learned indicator-function exponents) are obtained by minimizing the variational residual loss (eq. 9) or the SUPG-augmented loss (eq. 11), while the reported accuracy is the discrete L2 error (eq. 14) on a separate 100x100 equidistant grid. No equation defines one in terms of the other, so the improvements are not tautological. Hyperparameters tau, tau_g, and lambda are selected by grid search against the known solutions, but this is disclosed explicitly ('This search is usually not feasible in practice since the solution of the problem is usually not known') and the same oracle-style tuning is applied to the constant-parameter baselines, so the comparisons are symmetric rather than a hidden fit dressed as prediction. The self-citations to FastVPINNs (Anandh et al. 2024a,b; Ghose et al. 2024) provide the computational framework but are not the load-bearing justification for the accuracy claims, which are checked against external baselines (Sikora et al. 2023; Frerichs-Mihov et al. 2024a). The concern that the 100x100 grid cannot sample O(epsilon) layers is a question about whether the metric measures the intended quantity, i.e., a correctness/validity issue, not a circularity in the derivation chain.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the standard variational formulation of the CDR problem, the hard-constraint ansatz, and three heuristic assumptions: that the SUPG term is an effective regularizer in a neural loss, that minimizing the combined loss drives the network to the exact solution, and that the two manufactured benchmarks are representative. Four hyperparameters (tau, lambda, tau_g, kappa) are fit or hand-set using knowledge of the test problem, which weakens the external validity of the reported error reductions.

free parameters (4)
  • SUPG stabilization parameter tau (constant) = P_out: 3.5e-2; P_para: 1.2
    Grid-searched in Section 4.3 over ranges [1e-5, 1e-1] and [1e-5, 5] using the known exact solution to select the lowest L2 error; this fits the parameter to the test problem.
  • L2 weight regularization lambda = P_out: 2.4e-4; P_para: 4e-6
    Grid-searched in Section 4.3 over ranges [1e-5, 1e-3] and [1e-8, 1e-2] using the exact solution; only used in the regularized baseline, not in the main novelty.
  • tau growth factor tau_g = 1 for both problems
    Scales the network-predicted tau in Section 5.1; tuned over ranges [5e-2, 10] (P_out) and [1e-4, 10] (P_para) using the test problems, so the learned-tau result depends on a fitted hyperparameter.
  • Fixed indicator exponents kappa1, kappa2 = kappa1=30, kappa2=1e9
    Hand-chosen in equations (16) and (17) based on epsilon to set the slope of the hard-constraint indicator; used in all constant-parameter and learned-tau studies.
assumptions (5)
  • standard math Standard well-posedness and regularity of the CDR problem (1) with Lipschitz domain and bounded coefficients.
    Invoked in Section 2 for the variational formulation; standard in FEM analysis.
  • standard math The neural network ansatz u_hard = j + h u_NN with h satisfying (7) enforces the Dirichlet condition exactly.
    Used in Section 2, equation (6); standard hard-constraint construction from Lu et al. (2021b).
  • ad hoc to paper The SUPG term (10) is an appropriate regularizer for the variational loss.
    No analysis is provided for the neural-network setting; it is motivated by FEM theory and treated as a heuristic loss term.
  • ad hoc to paper Minimizing the combined loss (11) drives the network to the exact solution.
    Assumed implicitly; no convergence or landscape analysis is given.
  • ad hoc to paper The two benchmark problems with known manufactured solutions are representative of convection-dominated CDR problems.
    Underlies the generality of the conclusion in Section 6; only two problems, both with axis-aligned layers.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving hp-Variational Physics-Informed Neural Networks for Steady-State Convection-Dominated Problems." pith.science (2026). https://pith.science/paper/WDDN5RHM

@misc{pith2026241109329,
  author       = {Pith},
  title        = {Pith review of: Improving hp-Variational Physics-Informed Neural Networks for Steady-State Convection-Dominated Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WDDN5RHM}},
  note         = {Machine review of arXiv:2411.09329}
}
read the original abstract

This paper proposes and studies two extensions of applying hp-variational physics-informed neural networks, more precisely the FastVPINNs framework, to convection-dominated convection-diffusion-reaction problems. First, a term in the spirit of a SUPG stabilization is included in the loss functional and a network architecture is proposed that predicts spatially varying stabilization parameters. Having observed that the selection of the indicator function in hard-constrained Dirichlet boundary conditions has a big impact on the accuracy of the computed solutions, the second novelty is the proposal of a network architecture that learns good parameters for a class of indicator functions. Numerical studies show that both proposals lead to noticeably more accurate results than approaches that can be found in the literature.

Figures

Figures reproduced from arXiv: 2411.09329 by the authors.

Figure 1
Figure 1. Tensor-based loss computation schematic for FastVPINNs [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. hp-VPINNs architecture for convection-dominated problems with SUPG stabi￾lization, prescribed stabilization parameter, and hard constraints Subsequently, we address two convection-dominated problems, assessing the performance of the proposed loss functionals (11) and (12) with constant stabilization and regularization parameter, respectively, within the framework of hp-VPINNs and we compare the results and efficienc… view at source ↗
Figure 3
Figure 3. Exact Solution of PEJ for (a) ε = 0.1 , (b) ε = 0.01 and (c) ε = 0.001 Dirichlet conditions were imposed on ∂Ω. Solutions for various values of the diffusion coefficient are displayed in [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Exact solution for (a) Pout and (b) Ppara dominated convection-diffusion-reaction problems that satisfy Discrete Maximum Princi￾ples. 4.2 Validation of the FastVPINNs Implementation and Comparison with the Literature In this section, we consider the Eriksson–Johnson pr…
Figure 5
Figure 5. Figure 5: Best results for Pout (top) and Ppara (bottom). The exact solutions for Pout and Ppara are shown in (a) and (d). The predicted solutions are shown in (b) and (e) and the point-wise errors in (c) and (f) [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Sensitivity analysis on the τ parameter for problems Pout and Ppara using L SUPG τ as loss functional best one. In the case of Ppara, there is even just a single stabilization parameter that gives the by far most accurate result. The best values from [PITH_FULL_IMAGE:…
Figure 7
Figure 7. Figure 7: hp-VPINNs architecture for convection-dominated problems with SUPG stabi￾lization that predicts the stabilization parameter τ and imposes hard constraints 5 Learning the Parameters in the SUPG Loss and the Indicator Function The studies from Section 4 and further preli…
Figure 8
Figure 8. Figure 8: Results for problem Pout with L SUPG τ for τg = 1, 0.1, 0.05. Note that the optimal solution was obtained for τg = 1. The stabilization parameter τ (x, y) is shown in (a), (d) and (g). The predicted solution is depicted in (b), (e) and (h). The point-wise error is pres…
Figure 9
Figure 9. Figure 9: Results for problem Ppara with L SUPG τ for τg = 1, 0.1, 0.0005. Note that the optimal solution was obtained for τg = 1. The stabilization parameter τ (x, y) is shown in (a), (d) and (g). The predicted solution is depicted in (b), (e) and (h). The point-wise error is p…
Figure 10
Figure 10. Figure 10: Results for problem Pout for different ε. The indicator function h(x, y) is shown in (a), (d) and (g), the predicted solution in (b), (e) and (h), and the point-wise error in (c), (f) and (i) [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]
Figure 11
Figure 11. Figure 11: Results for problem Ppara for different ε. The indicator function h(x, y) is shown in (a), (d) and (g), the predicted solution in (b), (e) and (h), and the point-wise error in (c), (f) and (i) [PITH_FULL_IMAGE:figures/full_fig_p021_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 29 canonical work pages

  1. [1]

    Mart\' i n Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dandelion Man\' e , Rajat Monga, Sherry Moore, Der...

    2015

  2. [2]

    Abueidda, Qiyue Lu, and Seid Koric

    Diab W. Abueidda, Qiyue Lu, and Seid Koric. Meshless physics-informed deep learning method for three-dimensional solid mechanics. Internat. J. Numer. Methods Engrg., 122 0 (23): 0 7182--7201, 2021. ISSN 0029-5981,1097-0207

    work page 2021

  3. [3]

    FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks

    Thivin Anandh, Divij Ghose, and Sashikumaar Ganesan. FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks . Journal of Open Source Software, 9 0 (99): 0 6764, 2024 a

    work page 2024

  4. [4]

    FastVPINNs: Tensor-Driven Acceleration of VPINNs for Complex Geometries

    Thivin Anandh, Divij Ghose, Himanshu Jain, and Sashikumaar Ganesan. FastVPINNs: Tensor-Driven Acceleration of VPINNs for Complex Geometries . arXiv preprint arXiv:2404.12063, 2024 b

    arXiv 2024

  5. [5]

    Workshop report on basic research needs for scientific machine learning: Core technologies for artificial intelligence

    Nathan Baker, Frank Alexander, Timo Bremer, Aric Hagberg, Yannis Kevrekidis, Habib Najm, Manish Parashar, Abani Patra, James Sethian, Stefan Wild, Karen Willcox, and Steven Lee. Workshop report on basic research needs for scientific machine learning: Core technologies for artificial intelligence. USDOE Office of Science (SC), Washington, D.C. (United Stat...

    arXiv 2019

  6. [6]

    Barrenechea, Volker John, and Petr Knobloch

    Gabriel R. Barrenechea, Volker John, and Petr Knobloch. Finite element methods respecting the discrete maximum principle for convection-diffusion equations. SIAM Rev., 66 0 (1): 0 3--88, 2024. ISSN 0036-1445,1095-7200

    work page 2024

  7. [7]

    Brooks and Thomas J

    Alexander N. Brooks and Thomas J. R. Hughes. Streamline upwind/ P etrov- G alerkin formulations for convection dominated flows with particular emphasis on the incompressible N avier- S tokes equations. Comput. Methods Appl. Mech. Engrg., 32 0 (1-3): 0 199--259, 1982. ISSN 0045-7825. FENOMECH '81, Part I (Stuttgart, 1981)

    work page 1982

  8. [8]

    Scientific machine learning through physics-informed neural networks: where we are and what's next

    Salvatore Cuomo, Vincenzo Schiano Di Cola, Fabio Giampaolo, Gianluigi Rozza, Maziar Raissi, and Francesco Piccialli. Scientific machine learning through physics-informed neural networks: where we are and what's next. J. Sci. Comput., 92 0 (3): 0 Paper No. 88, 62 pages, 2022. ISSN 0885-7474,1573-7691

    work page 2022

Show all 33 references
  1. [9]

    Adaptive finite element methods for parabolic problems

    Kenneth Eriksson and Claes Johnson. Adaptive finite element methods for parabolic problems. I : A linear model problem. SIAM J. Numer. Anal., 28 0 (1): 0 43--77, 1991. ISSN 0036-1429

    1991

  2. [10]

    On loss functionals for physics-informed neural networks for steady-state convection-dominated convection-diffusion problems

    Derk Frerichs-Mihov, Linus Henning, and Volker John. On loss functionals for physics-informed neural networks for steady-state convection-dominated convection-diffusion problems. Comm. Appl. Math. Comp., page to appear, 2024 a

    2024

  3. [11]

    On collocation points for physics-informed neural networks applied to convection-dominated convection-diffusion problems

    Derk Frerichs-Mihov, Volker John, and Marwa Zainelabdeen. On collocation points for physics-informed neural networks applied to convection-dominated convection-diffusion problems. In Proceedings of ENUMATH 2023. Springer, 2024 b . accepted

    2023

  4. [12]

    Finite elements

    Sashikumaar Ganesan and Lutz Tobiska. Finite elements. Cambridge-IISc Series. Cambridge University Press, Cambridge, 2017. ISBN 978-1-108-41570-5. Theory and algorithms

    2017

  5. [13]

    FastVPINNs: A fast, versatile and robust Variational PINNs framework for forward and inverse problems in science

    Divij Ghose, Thivin Anandh, and Sashikumaar Ganesan. FastVPINNs: A fast, versatile and robust Variational PINNs framework for forward and inverse problems in science . In ICLR 2024 Workshop on Differential Equations in Science, 2024

    2024

  6. [14]

    T. J. R. Hughes and A. Brooks. A multidimensional upwind scheme with no crosswind diffusion. In Finite element methods for convection dominated flows ( P apers, W inter A nn. M eeting A mer. S oc. M ech. E ngrs., N ew Y ork, 1979) , volume 34 of AMD, pages 19--35. Amer. Soc. M...

    1979

  7. [15]

    V. John, J. M. Maubach, and L. Tobiska. Nonconforming streamline-diffusion-finite-element-methods for convection-diffusion problems. Numer. Math., 78 0 (2): 0 165--188, 1997. ISSN 0029-599X

    1997

  8. [16]

    On spurious oscillations at layers diminishing ( SOLD ) methods for convection-diffusion equations

    Volker John and Petr Knobloch. On spurious oscillations at layers diminishing ( SOLD ) methods for convection-diffusion equations. I . A review. Comput. Methods Appl. Mech. Engrg., 196 0 (17-20): 0 2197--2215, 2007. ISSN 0045-7825

    2007

  9. [17]

    On algebraically stabilized schemes for convection-diffusion-reaction problems

    Volker John and Petr Knobloch. On algebraically stabilized schemes for convection-diffusion-reaction problems. Numer. Math., 152 0 (3): 0 553--585, 2022. ISSN 0029-599X,0945-3245

    2022

  10. [18]

    Volker John, Petr Knobloch, and Simona B. Savescu. A posteriori optimization of parameters in stabilized methods for convection-diffusion problems--- P art I . Comput. Methods Appl. Mech. Engrg., 200 0 (41-44): 0 2916--2929, 2011. ISSN 0045-7825

    2011

  11. [19]

    A posteriori optimization of parameters in stabilized methods for convection-diffusion problems--- P art II

    Volker John, Petr Knobloch, and Ulrich Wilbrandt. A posteriori optimization of parameters in stabilized methods for convection-diffusion problems--- P art II . J. Comput. Appl. Math., 428: 0 Paper No. 115167, 17 pages, 2023. ISSN 0377-0427,1879-1778

    2023

  12. [20]

    Variational physics-informed neural networks for solving partial differential equations

    Ehsan Kharazmi, Zhongqiang Zhang, and George Em Karniadakis. Variational physics-informed neural networks for solving partial differential equations . arXiv preprint arXiv:1912.00873, 2019

    1912 arXiv

  13. [21]

    Ehsan Kharazmi, Zhongqiang Zhang, and George E. M. Karniadakis. hp - VPINN s: variational physics-informed neural networks with domain decomposition. Comput. Methods Appl. Mech. Engrg., 374: 0 Paper No. 113547, 25 pages, 2021. ISSN 0045-7825,1879-2138

    2021

  14. [22]

    Varnet: Variational neural networks for the solution of partial differential equations

    Reza Khodayi-Mehr and Michael Zavlanos. Varnet: Variational neural networks for the solution of partial differential equations. In Learning for Dynamics and Control, pages 298--307. PMLR, 2020

    2020

  15. [23]

    Characterizing possible failure modes in physics-informed neural networks

    Aditi Krishnapriyan, Amir Gholami, Shandian Zhe, Robert Kirby, and Michael W Mahoney. Characterizing possible failure modes in physics-informed neural networks. In Advances in Neural Information Processing Systems , volume 34, pages 26548--26560, Virtual, Online , 2021. Curran...

    2021

  16. [24]

    Lagaris, A

    I.E. Lagaris, A. Likas, and D.I. Fotiadis. Artificial neural networks for solving ordinary and partial differential equations . IEEE Transactions on Neural Networks, 9 0 (5): 0 987--1000, 1998

    1998

  17. [25]

    Deep XDE : a deep learning library for solving differential equations

    Lu Lu, Xuhui Meng, Zhiping Mao, and George Em Karniadakis. Deep XDE : a deep learning library for solving differential equations. SIAM Rev., 63 0 (1): 0 208--228, 2021 a . ISSN 0036-1445,1095-7200

    2021

  18. [26]

    Lu Lu, Rapha\"el Pestourie, Wenjie Yao, Zhicheng Wang, Francesc Verdugo, and Steven G. Johnson. Physics-informed neural networks with hard constraints for inverse design. SIAM J. Sci. Comput., 43 0 (6): 0 B1105--B1132, 2021 b . ISSN 1064-8275,1095-7197

    2021

  19. [27]

    Bound-preserving PINN s for steady-state convection-diffusion-reaction problems

    Marina Matthaiou, Volker John, and Marwa Zainelabdeen. Bound-preserving PINN s for steady-state convection-diffusion-reaction problems. WIAS Preprint No. 3134, 2024

    2024

  20. [28]

    https://developer.nvidia.com/modulus

    NVIDIA Modulus, 2024. https://developer.nvidia.com/modulus. Last accessed Jan 01, 2024

    2024

  21. [29]

    Psaros, Xuhui Meng, Zongren Zou, Ling Guo, and George Em Karniadakis

    Apostolos F. Psaros, Xuhui Meng, Zongren Zou, Ling Guo, and George Em Karniadakis. Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons. J. Comput. Phys., 477: 0 Paper No. 111902, 83 pages, 2023. ISSN 0021-9991,1090-2716

    2023

  22. [30]

    Raissi, P

    M. Raissi, P. Perdikaris, and G. E. Karniadakis. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys., 378: 0 686--707, 2019. ISSN 0021-9991,1090-2716

    2019

  23. [31]

    Robust numerical methods for singularly perturbed differential equations, volume 24 of Springer Series in Computational Mathematics

    Hans-G \"o rg Roos, Martin Stynes, and Lutz Tobiska. Robust numerical methods for singularly perturbed differential equations, volume 24 of Springer Series in Computational Mathematics. Springer-Verlag, Berlin, second edition, 2008. ISBN 978-3-540-34466-7. Convection-diffusion...

    2008

  24. [32]

    Physics informed neural networks with strong and weak residuals for advection-dominated diffusion problems, 2023

    Maciej Sikora, Patryk Krukowski, Anna Paszynska, and Maciej Paszynski. Physics informed neural networks with strong and weak residuals for advection-dominated diffusion problems, 2023

    2023

  25. [33]

    Artificial neural network-augmented stabilized finite element method

    Sangeeta Yadav and Sashikumaar Ganesan. Artificial neural network-augmented stabilized finite element method . J. Comput. Phys., 499: 0 Paper No. 112702, 15 pages, 2024. ISSN 0021-9991,1090-2716

    2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.