Pith. sign in

REVIEW 4 major objections 4 minor 44 references

Nonlinear model reduction with Neural Galerkin schemes on quadratic manifolds

T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper proposes reduced models that run Neural Galerkin dynamics on quadratic manifolds, giving locally unique, residual-minimizing trajectories and, for linear problems, online costs independent of the full-model state dimension.

desk verdict A solid, honest nonlinear-reduction paper whose core theorem is plausible; the online-efficiency claim rests on a tensor section with an apparent typo that needs fixing and verification. read the letter →

arxiv 2412.17695 v1 pith:7Q4HS45R submitted 2024-12-23 math.NA cs.NA

Philipp Weder , Paul Schwerdtner , Benjamin Peherstorfer This is my paper · ORCID
classification math.NAcs.NA MSC 65M6065M99
keywords nonlinearmodelreductionNeuralGalerkinschemesquadraticmanifoldsKolmogorovbarrieronlineefficiencyhyper-reductionresidualminimizationtransport-dominatedproblems
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 proposes reduced models that run Neural Galerkin dynamics on quadratic manifolds, a nonlinear parametrization built from a linear basis plus a quadratic feature map. The authors claim that these reduced solutions are locally unique and that, by construction, they minimize the residual norm at every time step, which keeps predictions stable even when the manifold is tightly fitted to training data. For linear full models, all quantities that depend on the full state dimension are moved into an offline phase, so each online step costs O($n^{4}$) with n the reduced dimension, independent of the full-model size N. For nonlinear problems, the residual can be evaluated at collocation points that are separate from the full-model grid, giving a built-in form of hyper-reduction without an extra empirical-interpolation layer. Numerical experiments with acoustic waves and charged-particle densities report orders-of-magnitude speedups over the full models.

What carries the argument

The machinery is the quadratic decoder $g_{V,W}(\theta) = s_0 + V\theta + W h(\theta)$ with $h(\theta) = \theta \otimes \theta$, whose Jacobian is affine in $\theta$: $J(\theta) = V + K\cdot\theta$. Neural Galerkin schemes choose the time derivative $\dot{\theta}$ by the least-squares projection of the residual onto the columns of this Jacobian, which is equivalent to minimizing the residual norm. The orthogonality $V^\top W = 0$ inherited from the greedy training algorithm makes $J(\theta)$ full column rank for all $\theta$ and lets the normal-equation matrix simplify to $I + K(\theta)^\top K(\theta)$, the key step in the local-uniqueness proof. For linear problems, tensor contractions precompute all $N$-dependent products offline, leaving only reduced-dimension operations in the online phase.

What would settle it

Train a quadratic manifold by ordinary least-squares fitting of $W$ without enforcing $V^\top W = 0$, or use collocation points distinct from grid points, then run QMNG on a linear wave problem and compute the smallest singular value of $J(\theta)$ along the trajectory; a zero singular value and a nonunique least-squares minimizer would refute the local-uniqueness claim outside the greedy setting.

Watch

Extended reading notes

Core claim

The central claim is that quadratic manifolds are not just expressive enough to approximate transport-dominated solutions; they also carry enough structure for Neural Galerkin schemes to define a well-posed reduced dynamics. The paper shows that when the manifold is trained with the greedy algorithm, the linear and quadratic parts satisfy $V^\top W = 0$, which forces the $N \times n$ Jacobian $J(\theta)$ to have full column rank for every $\theta$. Consequently, the least-squares problem that selects $\dot{\theta}$ has a unique solution, and the right-hand side of the reduced ODE is locally Lipschitz, so trajectories are locally unique. Because Neural Galerkin dynamics solve that least-squares problem, the reduced solution minimizes the residual norm over time; experiments on the acoustic wave and Vlasov problems show this residual-minimizing choice remains stable with tightly fitted manifolds where a constant-in-time test space fails. For linear full models, the same affine-Jacobian structure lets every $N$-dependent term be contracted in advance, so prediction costs scale as $O(n^4)$, independent of $N$; for nonlinear problems, the paper demonstrates that using roughly ten times fewer collocation points than grid points still tracks the reconstruction error.

Load-bearing premise

The uniqueness guarantee rests on training the quadratic manifold with the paper's greedy algorithm so that the linear and quadratic parts are exactly orthogonal; if that orthogonality is lost, the Jacobian can lose rank and the reduced trajectory may not be unique.

Editorial extensions

If this is right

  • Quadratic-manifold Neural Galerkin reduced models have locally unique trajectories, so they do not suffer the tangent-space collapse that can occur with other nonlinear parametrizations.
  • For linear full models, online predictions of QMNG reduced models cost $O(n^4)$ per step independently of the full-model dimension $N$, which is what the paper means by online efficiency.
  • Because collocation points can differ from grid points, nonlinear QMNG reduced models carry hyper-reduction within the dynamics and need no separate empirical-interpolation step.
  • Residual-norm minimization is the ingredient that stabilizes QMNG predictions on tightly fitted manifolds; replacing the Jacobian by a constant-in-time test space destabilizes the reduced dynamics, as the acoustic-wave and charged-particle experiments show.

Reading between the lines

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

  • The local-uniqueness guarantee is tied to the greedy training property $V^\top W = 0$; the paper notes this property can fail once interpolation separates collocation points from grid points, so users of the interpolated variant should verify the rank of the batch Jacobian before relying on uniqueness.
  • The $O(n^4)$ online cost means the practical speedup window is limited in the reduced dimension: the acoustic-wave experiment itself shows speedup disappears around $n = 40$, so the method will be most useful for small $n$ with large $N$.
  • For nonlinear problems, the paper reports no speedup on the Burgers example because the full model is cheap and automatic differentiation dominates; extending the precomputation idea to nonlinear right-hand sides, or approximating derivatives without automatic differentiation, would be the natural next step.
  • The stability comparison suggests a direct test of the time-continuous versus time-discrete residual formulation on the same quadratic manifolds; if QMNG remains stable where discrete-residual fits fail, the paper's variational choice is validated as the cause.
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 combines Neural Galerkin schemes with quadratic-manifold parametrizations for nonlinear model reduction. For the vector-valued decoder with collocation points equal to the full-model grid, the authors prove local uniqueness of the reduced dynamics and residual-norm minimization, and for linear full models they derive precomputable tensor quantities that purportedly make online predictions cost O(n^4), independent of the full state dimension N. A second variant uses an interpolated parametrization that allows collocation points to differ from the grid points, which acts as a form of hyper-reduction. The method is tested on advecting acoustic waves, charged-particle densities, and Burgers' equation, with honest reporting of where speedups are and are not obtained.

Significance. If the main technical claims hold, the paper makes a useful contribution: residual-norm minimization distinguishes QMNG reduced models from constant-test-space approaches and appears to stabilize predictions on tightly fitted quadratic manifolds, while the O(n^4) online complexity for linear problems is a structurally interesting result. The paper is also careful in stating assumptions and limitations, including the absence of speedup for Burgers' equation and the dependence of the uniqueness result on the greedy-manifold property V^T W = 0. However, the central tensor precomputation in Section 4.3 contains a visible indexing/definition error and is never derived, so the online-efficiency claim and the linear-problem experiments are not yet fully supported.

major comments (4)
  1. [Section 4.3, Eq. (4.10)] The definition of H_{kjm} is printed identically to A_{kjm}, namely H_{kjm} = K_{ijk} A_{il} V_{lm}, so the last index is m (of size n) even though H is used in (4.12) to multiply θ ⊗ θ, which has length n^2. Moreover the right-hand side must involve A W, not A V, to represent the term K(θ)^T A W (θ ⊗ θ). As printed, the definition of H cannot be correct, and since Algorithm 3 assembles this tensor for the linear-problem experiments, the O(n^4)-independent-of-N claim is not currently verified.
  2. [Section 4.3, Eqs. (4.11)-(4.12)] Please provide a term-by-term derivation of (4.12) from the normal equations of (4.5). The displayed formula does not follow as written: with f(q)=Aq and q=s0+Vθ+W(θ⊗θ), the first projection term is V^T A s0, not the V^T s0 defined in (4.9), and the term θ·S with S_{jk}=K_{ikj}[s0]_i cannot by itself account for A s0. Either the tensor definitions or the formula for fhat need to be corrected, and the derivation should be included so that the precomputation can be checked.
  3. [Sections 3.4 and 4.3, tensor-index conventions] The tensor products in (3.11)-(3.12) are defined for T ∈ R^{n×m×p}, but (3.13) declares KΞ ∈ R^{m×n×n}; under the stated convention the expression KΞ·θ is not well defined. Likewise the remark below (4.8) that θ·J·θ = (θ·K)^T(K·θ) mixes the two tensor-vector products in a way that is dimensionally inconsistent. Please make all tensor index conventions explicit and consistent throughout Section 4.3, since the complexity table and Algorithms 3-4 rely on these operations.
  4. [Abstract, Section 4.2, Section 6] The abstract and conclusions state that solutions of the proposed QMNG reduced models are locally unique without the qualifications required by Lemma 4.1 and Proposition 4.2. Section 4.2 itself notes that the condition V_Ξ^T W_Ξ = 0 can be violated when interpolation separates collocation points from grid points, so the uniqueness guarantee does not cover the interpolated QMNG models used for the Burgers experiment. Please qualify the claims to state that local uniqueness holds for the grid-collocated vector-valued version under the greedy-manifold construction with V^T W = 0.
minor comments (4)
  1. [Section 5.1.2] The text says the QMNG approximations are plotted for t ∈ {0,2,4,6.8}, while Figure 1 uses t ∈ {0,2,4,6,8}; the former appears to be a typo.
  2. [Algorithm 2] The procedure name is printed as OMNGOnline; this should be QMNGOnline.
  3. [Eq. (4.3)] The Jacobian matrix is said to be in R^{N×n}, but the displayed entries use ∂/∂θ_p; p is undefined and should presumably be n.
  4. [Section 4.3] The paper states that solving (4.5) via normal equations uses J^T J = I + K^T K, which is plausible under V^T W = 0, but the corresponding four-tensor J in (4.8) is never explicitly constructed. A short construction of J from K would improve verifiability.

Circularity Check

1 steps flagged · score 2.0 of 10

Residual-minimization claim restates the least-squares definition; substantive uniqueness and online-efficiency results are not circular.

  1. self definitional [Section 3.5 (after Eq. (3.5)); also Abstract]
    "The QMNG reduced model sets the residual at θ(t; µ) orthogonal to the space spanned by the columns of the Jacobian (3.8) at θ(t; µ), which is equivalent to minimizing the residual in the Euclidean norm over the collocation points ξ1, . . . ,ξm as in the least-squares problem (3.5). Thus, solutions of QMNG reduced models minimize the residual norm over time t in the sense of (3.5)."

    The QMNG reduced model is defined by the least-squares problem (3.5), whose objective is exactly the squared residual norm. The claim that solutions minimize the residual norm is therefore a direct restatement of the definition, not an independent derivation. The abstract presents this as a shown property ('minimize the residual norm over time'), but it holds by construction. This definitional step is minor: the local-uniqueness result (Prop. 4.2) and the online-efficiency claim (Section 4.3) are derived from the model equations and the algebraic condition V^T W=0, not fitted to test data.

full rationale

The main reduced dynamics are obtained by solving a residual least-squares problem (3.5)/(4.5), not by fitting test trajectories; the numerical comparisons use held-out parameters, so the accuracy and speedup results are not forced by construction. The local-uniqueness result (Prop. 4.2) is a genuine consequence of Lemma 4.1, which uses the property V^T W=0; this property is attributed to the authors' own greedy construction [36] and also to [9], but it is an algebraic condition on the trained manifold rather than a restatement of the uniqueness theorem, and the paper explicitly records where it fails (interpolated V_Ξ, Section 4.2). Thus the reliance on [36] is a self-citation, not a circular reduction. The online-efficiency claim (O(n^4) independent of N) follows from precomputing N-dependent tensors and is an algorithmic complexity statement; whether Eq. (4.10)-(4.12) are implemented correctly (e.g., the printed H definition identical to A) is a correctness risk, not circularity. The only definitional element is the residual-minimization statement in Section 3.5, which restates the least-squares definition of the QMNG dynamics; this minor self-definitional framing does not undermine the paper's substantive derivation.

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

The central results rest on the greedy-manifold property V^T W = 0 and on standard smoothness and discretization assumptions. The methods introduce no new physical entities.

free parameters (3)
  • gamma (regularization in W fit) = 1e-6 (all three experiments)
    Manually selected based on training data; controls the fit of the quadratic manifold. Varying it changes error and stability (Figs. 3, 7).
  • l (candidate singular vectors) = 200
    Hyperparameter of the greedy manifold construction [36].
  • m (number of collocation points) = 256, 512, 1024 for Burgers; N for linear problems
    Choice of collocation points affects accuracy (Fig. 9b).
assumptions (4)
  • domain assumption Greedy quadratic-manifold construction yields V^T W = 0
    Used in Lemma 4.1 and Prop 4.2 to prove full column rank of the Jacobian and local uniqueness. If another construction is used, the guarantee may fail (as the paper notes for interpolated V_Ξ^T W_Ξ).
  • standard math Right-hand side f(·, mu) is continuously differentiable
    Needed for Proposition 4.2 to apply Picard-Lindelöf; standard for the semi-discrete ODEs considered.
  • domain assumption Interpolation operator I is accurate enough
    The interpolated decoder g_{V,W,I} must approximate the underlying functions at the collocation points; accuracy is assumed in Section 3.2.
  • domain assumption Collocation points approximate the L2 inner product
    The least-squares problem (3.5) uses the Euclidean norm over Xi as a proxy for the L2 norm over Omega, which is standard but an approximation (Section 3.3).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nonlinear model reduction with Neural Galerkin schemes on quadratic manifolds." pith.science (2026). https://pith.science/paper/7Q4HS45R

@misc{pith2026241217695,
  author       = {Pith},
  title        = {Pith review of: Nonlinear model reduction with Neural Galerkin schemes on quadratic manifolds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7Q4HS45R}},
  note         = {Machine review of arXiv:2412.17695}
}
read the original abstract

Leveraging nonlinear parametrizations for model reduction can overcome the Kolmogorov barrier that affects transport-dominated problems. In this work, we build on the reduced dynamics given by Neural Galerkin schemes and propose to parametrize the corresponding reduced solutions on quadratic manifolds. We show that the solutions of the proposed quadratic-manifold Neural Galerkin reduced models are locally unique and minimize the residual norm over time, which promotes stability and accuracy. For linear problems, quadratic-manifold Neural Galerkin reduced models achieve online efficiency in the sense that the costs of predictions scale independently of the state dimension of the underlying full model. For nonlinear problems, we show that Neural Galerkin schemes allow using separate collocation points for evaluating the residual function from the full-model grid points, which can be seen as a form of hyper-reduction. Numerical experiments with advecting waves and densities of charged particles in an electric field show that quadratic-manifold Neural Galerkin reduced models lead to orders of magnitude speedups compared to full models.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 39 canonical work pages

  1. [9]

    Geelen, S

    R. Geelen, S. Wright, K. Willcox, Operator inference for non-intrusive model reduction with quadratic manifolds, Comput. Methods Appl. Mech. Engrg. 403 (2023) 115717

    work page 2023

  2. [33]

    Sharma, H

    H. Sharma, H. Mu, P. Buchfink, R. Geelen, S. Glas, B. Kramer, Symplectic model reduction of hamiltonian systems using data-driven quadratic manifolds, Computer Methods in Applied Mechanics and Engineering 417 (2023) 116402

    2023

  3. [36]

    Schwerdtner, B

    P. Schwerdtner, B. Peherstorfer, Greedy construction of quadratic manifolds for nonlinear dimen- sionality reduction and nonlinear model reduction, arXiv 2403.06732 (2024)

    arXiv 2024

  4. [1]

    Rozza, D

    G. Rozza, D. Huynh, A. Patera, Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations, Archives of Computational 22 0.0 2.5 5.0 full model t=0.0 t=0.2 t=0.5 t=0.8 t=1.0 0.0 2.5 5.0 QM recon. −1 0 1 x 0.0 2.5 5.0 QMNG −1 0 1 x −1 0 1 x −1 0 1 x −1 0 1 x Figure 8: Burgers’ equ...

    work page 2007

  5. [2]

    Benner, S

    P. Benner, S. Gugercin, K. Willcox, A survey of projection-based model reduction methods for parametric dynamical systems, SIAM Rev. 57 (4) (2015) 483–531

    work page 2015

  6. [3]

    A. C. Antoulas, C. A. Beattie, S. Gugercin, Interpolatory Methods for Model Reduction, SIAM, 2021

    work page 2021

  7. [4]

    Peherstorfer, Breaking the Kolmogorov barrier with nonlinear model reduction, Notices of the American Mathematical Society 69 (5) (2022) 725–733

    B. Peherstorfer, Breaking the Kolmogorov barrier with nonlinear model reduction, Notices of the American Mathematical Society 69 (5) (2022) 725–733

    work page 2022

  8. [5]

    Bruna, B

    J. Bruna, B. Peherstorfer, E. Vanden-Eijnden, Neural Galerkin schemes with active learning for high-dimensional evolution equations, Journal of Computational Physics 496 (2024) 112588

    work page 2024

Show all 44 references
  1. [6]

    Berman, P

    J. Berman, P. Schwerdtner, B. Peherstorfer, Chapter 8 - Neural Galerkin schemes for sequential- in-time solving of partial differential equations with deep networks, in: S. Mishra, A. Townsend (Eds.), Numerical Analysis Meets Machine Learning, Vol. 25 of Handbook of Numerical ...

    2024

  2. [7]

    Zhang, Y

    H. Zhang, Y. Chen, E. Vanden-Eijnden, B. Peherstorfer, Sequential-in-time training of nonlin- ear parametrizations for solving time-dependent partial differential equations, arXiv 2404.01145 (2024)

    2024 arXiv

  3. [8]

    S. Jain, P. Tiso, J. B. Rutzmoser, D. J. Rixen, A quadratic manifold for model order reduction of nonlinear structural dynamics, Computers & Structures 188 (2017) 80–94

    2017

  4. [10]

    Barnett, C

    J. Barnett, C. Farhat, Quadratic approximation manifold for mitigating the Kolmogorov barrier in nonlinear projection-based model order reduction, J. Comput. Phys. 464 (2022) 111348

    2022

  5. [11]

    Barrault, Y

    M. Barrault, Y. Maday, N. C. Nguyen, A. T. Patera, An ‘empirical interpolation’ method: appli- cation to efficient reduced-basis discretization of partial differential equations, C. R. Acad. Sci. Paris, Ser. I 339 (9) (2004) 667–672

    2004

  6. [12]

    P. A. M. Dirac, Note on exchange phenomena in the Thomas atom, Mathematical Proceedings of the Cambridge Philosophical Society 26 (3) (1930) 376–385

    1930

  7. [13]

    Frenkel, Wave Mechanics, Advanced General Theory, Clarendon Press, Oxford, 1934

    J. Frenkel, Wave Mechanics, Advanced General Theory, Clarendon Press, Oxford, 1934

    1934

  8. [14]

    Lubich, From quantum to classical molecular dynamics: reduced models and numerical anal- ysis, Vol

    C. Lubich, From quantum to classical molecular dynamics: reduced models and numerical anal- ysis, Vol. 12, European Mathematical Society, 2008

    2008

  9. [15]

    O. Koch, C. Lubich, Dynamical low-rank approximation, SIAM Journal on Matrix Analysis and Applications 29 (2) (2007) 434–454

    2007

  10. [16]

    Black, P

    F. Black, P. Schulze, B. Unger, Projection-based model reduction with dynamically transformed modes, ESAIM: M2AN 54 (6) (2020) 2011–2043

    2020

  11. [17]

    Y. Du, T. A. Zaki, Evolutional deep neural network, Physical Review E 104 (4) (oct 2021)

    2021

  12. [18]

    Anderson, M

    W. Anderson, M. Farazmand, Evolution of nonlinear reduced-order solutions for PDEs with conserved quantities, SIAM Journal on Scientific Computing 44 (1) (2022) A176–A197

    2022

  13. [19]

    M. Kast, J. S. Hesthaven, Positional embeddings for solving PDEs with evolutional deep neural networks, arXiv 2308.03461 (2023). 24

    2023 arXiv

  14. [20]

    Schulze, Structure-preserving model reduction for port-hamiltonian systems based on a special class of nonlinear approximation ansatzes, arXiv 2302.06479 (2023)

    P. Schulze, Structure-preserving model reduction for port-hamiltonian systems based on a special class of nonlinear approximation ansatzes, arXiv 2302.06479 (2023)

    2023 arXiv

  15. [21]

    Lasser, C

    C. Lasser, C. Lubich, Computing quantum dynamics in the semiclassical regime, Acta Numerica 29 (2020) 229–401

    2020

  16. [22]

    J. S. Hesthaven, C. Pagliantini, G. Rozza, Reduced basis methods for time-dependent problems, Acta Numer. 31 (2022) 265–345

    2022

  17. [23]

    K. Lee, K. T. Carlberg, Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders, J. Comput. Phys. 404 (2020) 108973

    2020

  18. [24]

    Y. Kim, Y. Choi, D. Widemann, T. Zohdi, A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder, Journal of Computational Physics 451 (2022) 110841

    2022

  19. [25]

    P. J. Blonigan, F. Rizzi, M. Howard, J. A. Fike, K. T. Carlberg, Model reduction for steady hy- personic aerodynamics via conservative manifold least-squares Petrov–Galerkin projection, AIAA Journal 59 (4) (2021) 1296–1312

    2021

  20. [26]

    Cocola, J

    J. Cocola, J. Tencer, F. Rizzi, E. Parish, P. Blonigan, Hyper-reduced autoencoders for efficient and accurate nonlinear model reductions, arXiv 2303.09630 (2023)

    2023 arXiv

  21. [27]

    Romor, G

    F. Romor, G. Stabile, G. Rozza, Non-linear manifold reduced-order models with convolutional autoencoders and reduced over-collocation method, Journal of Scientific Computing 94 (3) (2023) 74

    2023

  22. [28]

    Berman, B

    J. Berman, B. Peherstorfer, CoLoRA: Continuous low-rank adaptation for reduced implicit neu- ral modeling of parameterized partial differential equations, in: R. Salakhutdinov, Z. Kolter, K. Heller, A. Weller, N. Oliver, J. Scarlett, F. Berkenkamp (Eds.), Proceedings of the 41...

    2024

  23. [29]

    Berman, B

    J. Berman, B. Peherstorfer, Randomized sparse Neural Galerkin schemes for solving evolution equations with deep networks, in: A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, S. Levine (Eds.), Advances in Neural Information Processing Systems, Vol. 36, Curran Associates, ...

    2023

  24. [30]

    Rutzmoser, D

    J. Rutzmoser, D. Rixen, P. Tiso, S. Jain, Generalization of quadratic manifolds for reduced order modeling of nonlinear structural dynamics, Computers & Structures 192 (2017) 196–209

    2017

  25. [31]

    Geelen, L

    R. Geelen, L. Balzano, K. Willcox, Learning latent representations in high-dimensional state spaces using polynomial manifold constructions (2023)

    2023

  26. [32]

    Geelen, L

    R. Geelen, L. Balzano, S. Wright, K. Willcox, Learning physics-based reduced-order models from data using nonlinear manifolds (2023)

    2023

  27. [34]

    Cohen, C

    A. Cohen, C. Farhat, Y. Maday, A. Somacal, Nonlinear compressive reduced basis approximation for PDE’s, Comptes Rendus. M´ ecanique 351 (S1) (2023) 357–374

    2023

  28. [35]

    Benner, P

    P. Benner, P. Goyal, J. Heiland, I. Pontes Duff, A quadratic decoder approach to nonintrusive reduced-order modeling of nonlinear dynamical systems, PAMM 23 (1) (2023) e202200049. 25

    2023

  29. [37]

    Schwerdtner, P

    P. Schwerdtner, P. Mohan, A. Pachalieva, J. Bessac, D. O’Malley, B. Peherstorfer, Online learning of quadratic manifolds from streaming data for nonlinear dimensionality reduction and nonlinear model reduction, arXiv 2409.02703 (2024)

    2024 arXiv

  30. [38]

    Goyal, P

    P. Goyal, P. Benner, Generalized quadratic embeddings for nonlinear dynamics using deep learn- ing, Phys. D 463 (2024) 134158

    2024

  31. [39]

    Ballout, Y

    H. Ballout, Y. Maday, C. Prud’homme, Nonlinear compressive reduced basis approximation for multi-parameter elliptic problem, Springer Nature Switzerland, Cham, 2024, pp. 55–73

    2024

  32. [40]

    Schwerdtner, S

    P. Schwerdtner, S. Gugercin, B. Peherstorfer, Empirical sparse regression on quadratic manifolds, arXiv 2412.09746 (2024)

    2024 arXiv

  33. [41]

    Barnett, C

    J. Barnett, C. Farhat, Y. Maday, Neural-network-augmented projection-based model order reduc- tion for mitigating the Kolmogorov barrier to reducibility, J. Comput. Phys. 492 (2023) 112420

    2023

  34. [42]

    Y. Wen, E. Vanden-Eijnden, B. Peherstorfer, Coupling parameter and particle dynamics for adap- tive sampling in neural galerkin schemes, Physica D: Nonlinear Phenomena 462 (2024) 134129

    2024

  35. [43]

    T. G. Kolda, B. W. Bader, Tensor decompositions and applications, SIAM Review 51 (3) (2009) 455–500

    2009

  36. [44]

    Schwerdtner, P

    P. Schwerdtner, P. Schulze, J. Berman, B. Peherstorfer, Nonlinear embeddings for conserving Hamiltonians and other quantities with Neural Galerkin schemes, SIAM Journal on Scientific Computing 46 (5) (2024) C583–C607. 26

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.