Neural Network Approximation and Structural Complexity of Semilinear PDEs with Lower Bounds
Abstract
Neural networks are commonly used to learn the solution operators of PDEs such as the Navier-Stokes equations and viscous conservation laws. On standard benchmarks, moderately sized and often shallow networks achieve small errors that decay at algebraic rates with the network size. These observations are not well described by the current approximation theory, which treats the solution operator as a generic map and derives worst-case bounds over large classes of operators. We show that the structure of the equations and their properties explain the observed behavior. For semilinear parabolic equations, the approximation costs separate into two terms, where one is controlled by an invariant of the governing law and the other by the data. The constitutive law can be parameterized as a single block whose size for accuracy η is O(η−pχlogη−1) for every p>pf, where the structural exponent pf only depends on the computation graph of f. This exponent does not depend on the dimension and is in fact zero for polynomial laws, so for the two-dimensional Navier-Stokes equations the nonlinear block costs only O(logη−1) parameters. The data enters through encoding layers with a bounded number of learned weights per mode, so neither term grows with the grid resolution or the time step. While the analysis only uses the max-norm error of the pointwise block, we show that one obtains approximation bounds in a stronger norm along the whole trajectory. For viscous Hamilton-Jacobi equations with a Cs Hamiltonian, the approximation requires on the order of η−1/s parameters, and we prove a matching lower bound, up to logarithms. This shows that solution operators for this class of problems have low structural complexity, even though they are nonlinear maps between infinite-dimensional spaces.
Related
- Arbitrary-Accuracy Neural Approximation with Optimal Neuron Count and Near-Optimal Bit Complexity
- Neural network enhancement of the Born approximation
- ReLU-spline activation functions for Neural Network approximation of solutions to differential equations
- Shallow neural network approximation in mixed Sobolev spaces
- Universal neural network approximation of symmetric polyconvex energies and envelopes
BibTeX
@misc{schaeffer2026neural,
title = {{Neural Network Approximation and Structural Complexity of Semilinear PDEs with Lower Bounds}},
author = {Hayden Schaeffer},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.40384},
eprint = {2610.40384},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.40384},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Neural Network Approximation and Structural Complexity of Semilinear PDEs with Lower Bounds
AU - Hayden Schaeffer
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.40384
UR - https://arxiv.org/abs/2610.40384
AB - Neural networks are commonly used to learn the solution operators of PDEs such as the Navier-Stokes equations and viscous conservation laws. On standard benchmarks, moderately sized and often shallow networks achieve small errors that decay at algebraic rates with the network size. These observations are not well described by the current approximation theory, which treats the solution operator as a generic map and derives worst-case bounds over large classes of operators. We show that the structure of the equations and their properties explain the observed behavior. For semilinear parabolic equations, the approximation costs separate into two terms, where one is controlled by an invariant of the governing law and the other by the data. The constitutive law can be parameterized as a single block whose size for accuracy η is O(η^−pχlogη^−1) for every p>p_f, where the structural exponent p_f only depends on the computation graph of f. This exponent does not depend on the dimension and is in fact zero for polynomial laws, so for the two-dimensional Navier-Stokes equations the nonlinear block costs only O(logη^−1) parameters. The data enters through encoding layers with a bounded number of learned weights per mode, so neither term grows with the grid resolution or the time step. While the analysis only uses the max-norm error of the pointwise block, we show that one obtains approximation bounds in a stronger norm along the whole trajectory. For viscous Hamilton-Jacobi equations with a C^s Hamiltonian, the approximation requires on the order of η^−1/s parameters, and we prove a matching lower bound, up to logarithms. This shows that solution operators for this class of problems have low structural complexity, even though they are nonlinear maps between infinite-dimensional spaces.
ER -