GAIA introduces a geometry-adaptive integral autoencoder that unifies forward, boundary-value, and inverse PDE operator learning on arbitrary domains via geometry tokens and cross-attention.
hub Mixed citations
Neural Operator: Graph Kernel Network for Partial Differential Equations
Mixed citation behavior. Most common role is background (64%).
70
Pith papers citing it
Background
64% of classified citations
abstract
The classical development of neural networks has been primarily for mappings between a finite-dimensional Euclidean space and a set of classes, or between two finite-dimensional Euclidean spaces. The purpose of this work is to generalize neural networks so that they can learn mappings between infinite-dimensional spaces (operators). The key innovation in our work is that a single set of network parameters, within a carefully designed network architecture, may be used to describe mappings between infinite-dimensional spaces and between different finite-dimensional approximations of those spaces. We formulate approximation of the infinite-dimensional mapping by composing nonlinear activation functions and a class of integral operators. The kernel integration is computed by message passing on graph networks. This approach has substantial practical consequences which we will illustrate in the context of mappings between input data to partial differential equations (PDEs) and their solutions. In this context, such learned networks can generalize among different approximation methods for the PDE (such as finite difference or finite element methods) and among approximations corresponding to different underlying levels of resolution and discretization. Experiments confirm that the proposed graph kernel network does have the desired properties and show competitive performance compared to the state of the art solvers.
hub tools
citation-role summary
background 7
method 3
dataset 1
citation-polarity summary
representative citing papers
MEEC equips point clouds with a discrete exterior calculus that satisfies exact conservation and is differentiable in point positions, allowing a single trained kernel to produce compatible physics on unseen geometries and parameters.
Fourier Neural Operator parameterizes integral kernels in Fourier space to learn parametric PDE solution operators, delivering up to 1000x speedups and zero-shot super-resolution on turbulent Navier-Stokes flows.
A self-explainable operator learning method reformulates operators as decomposable integral equations to reveal spatial input contributions to predictions in blood flow and aerodynamics problems.
PNOT combines graph attention on boundary heat flux with a physics-aware neural operator and gradient-constrained loss to reconstruct divertor temperature fields for real-time fusion control.
HO-FNO extends standard FNO with n-linear spectral mixing and shows improved accuracy on nonlinear PDE benchmarks, sometimes with a single layer beating deeper FNO models.
TNOs lift neural operators to topological cell complexes via Discrete Exterior Calculus for cross-dimensional coupling, subsuming prior NOs and showing accuracy gains on PDE benchmarks with irregular geometries.
LiNO introduces a light-transport-inspired decomposition of neural operator latent dynamics into pointwise reflection/refraction and input-dependent scattering for improved structure in operator learning.
Cellular Sheaf Neural Operators use cell complexes, learned restriction maps, and structure-aware message passing to create discretization-aware neural surrogates that preserve constraints in multiphysics PDEs such as MHD.
Zero-shot super-resolution is information-theoretically impossible for some simple operators but possible under Hölder smoothness of outputs, accompanied by generalization bounds.
Functional Attention replaces pairwise softmax attention with structured linear operators inspired by geometric functional maps to produce compact, resolution-invariant representations for operator learning.
Neural networks parameterize finite-rank generators for ODEs on the orthogonal Lie group, allowing optimization of orthonormal bases in function space with a universality result that rank-2 generators suffice for density.
Fine-tuning neural PDE operators to regime endpoints reveals a physical direction in weight space that CCM uses to compose accurate merged models for new or extrapolated regimes from metadata or short prefixes.
Local neural operators on 3x3x3 patches, composed via Schwarz iteration, solve large-scale nonlinear elasticity on arbitrary geometries without domain-specific retraining.
Proposes a vector-valued RKHS framework for Bayesian optimization with structured measurements, deriving concentration bounds and UCB-based regret guarantees that recover sublinear rates.
QuadNorm uses quadrature-based moments instead of uniform averaging in normalization layers, achieving O(h²) consistency across resolutions and better cross-resolution transfer in neural operators.
HIN-LRI augments a low-regularity integrator with a latent-manifold neural correction trained end-to-end on trajectory error to improve accuracy on nonlinear dispersive equations with rough data.
A multilinear operator learned on PCA coefficients maps time-since-ignition inputs to smoke outputs, matching Monte Carlo accuracy with half the model calls and outperforming prior classifiers on holdout data.
Hybrid FNO-LBM accelerates porous media flow convergence by up to 70% via neural initialization and stabilizes unsteady simulations through embedded FNO rollouts, allowing small models to match larger ones in accuracy.
Physics-informed Fourier neural operators recover plasmoid formation in sparse SRRMHD vortex data where data-only models fail, and transformer operators approximate AMR jet evolution, marking first reported uses in these relativistic MHD settings.
Parameterizing the temporal derivative in PINNs and reconstructing via Volterra integral yields 100-200x lower errors on advection, Burgers, and Klein-Gordon equations while proving equivalence to the original PDE.
PGOT uses physics-geometry attention and spatially adaptive computation paths to achieve state-of-the-art results on complex PDE benchmarks and industrial geometry tasks.
WHNO using Walsh-Hadamard transforms outperforms Fourier Neural Operators on PDEs with discontinuous coefficients, and optimal WHNO-FNO ensembles cut mean-squared error by 35-40 percent.
SAE-NOs extend sparse autoencoders to function spaces via Fourier neural operators with concept and domain sparsity, learning localized patterns more efficiently and generalizing across discretizations on vision data.
citing papers explorer
-
Learning Neural Operator Surrogates for the Black Hole Accretion Code
Physics-informed Fourier neural operators recover plasmoid formation in sparse SRRMHD vortex data where data-only models fail, and transformer operators approximate AMR jet evolution, marking first reported uses in these relativistic MHD settings.
-
Recovering Physical Dynamics from Discrete Observations via Intrinsic Differential Consistency
Enforcing semi-group consistency on a time-conditioned secant velocity field via Symmetry Rupture improves rollout accuracy and efficiency when learning physical dynamics from discrete observations.
-
Universal Neural Propagator: Learning Time Evolution in Many-Body Quantum Systems
The Universal Neural Propagator is a single neural model trained self-supervised to predict time evolution in driven quantum many-body systems across arbitrary protocols and initial states.