Pith. sign in

REVIEW 23 cited by

Neural Operator: Learning Maps Between Function Spaces

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2108.08481 v6 pith:UZ7V6PQ6 submitted 2021-08-19 cs.LG cs.NAmath.NA

Nikola Kovachki , Zongyi Li , Burigede Liu , Kamyar Azizzadenesheli , Kaushik Bhattacharya , Andrew Stuart , Anima Anandkumar This is my paper
classification cs.LGcs.NAmath.NA
keywords neuraloperatorslearningoperatorspacesfunctionproposeddimensional
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

The classical development of neural networks has primarily focused on learning mappings between finite dimensional Euclidean spaces or finite sets. We propose a generalization of neural networks to learn operators, termed neural operators, that map between infinite dimensional function spaces. We formulate the neural operator as a composition of linear integral operators and nonlinear activation functions. We prove a universal approximation theorem for our proposed neural operator, showing that it can approximate any given nonlinear continuous operator. The proposed neural operators are also discretization-invariant, i.e., they share the same model parameters among different discretization of the underlying function spaces. Furthermore, we introduce four classes of efficient parameterization, viz., graph neural operators, multi-pole graph neural operators, low-rank neural operators, and Fourier neural operators. An important application for neural operators is learning surrogate maps for the solution operators of partial differential equations (PDEs). We consider standard PDEs such as the Burgers, Darcy subsurface flow, and the Navier-Stokes equations, and show that the proposed neural operators have superior performance compared to existing machine learning based methodologies, while being several orders of magnitude faster than conventional PDE solvers.

Discussion (0). Sign in to comment.

Forward citations

Cited by 23 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 79 citations worldwide. Full citation record

  1. Elliptic Regularity Theory in Barron Spaces and Applications to the Deep Ritz Method

    math.AP 2026-07 accept novelty 7.0 of 10

    Harmonic functions with Barron Dirichlet data fail to be Lipschitz or H², yet admit Barron approximants of norm ~|log ε| with error ~ε on half-spaces and 2D rectangles, giving Deep Ritz a priori rates.

  2. Quantitative Sobolev Approximation Bounds for Neural Operators with Empirical Validation on Burgers Equation

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    Neural operators approximate continuous operators from H^s to H^t with O(N^{-s/d}) error in H^t norm; FNOs on Burgers achieve H^1 errors to 10^{-7} and follow a power-law scaling with exponent ~1.4.

  3. Flow Straight and Fast in Hilbert Space: Functional Rectified Flow

    cs.LG 2025-09 conditional novelty 7.0 of 10

    Functional rectified flow is defined and proved to preserve marginals in separable Hilbert spaces, with functional flow matching and probability-flow ODEs as special cases.

  4. tFUSOperator: Operator Learning for Transcranial Focused Ultrasound Digital Twins

    cs.LG 2026-08 conditional novelty 6.0 of 10

    tFUSOperator predicts focused ultrasound pressure fields on seen and unseen skulls from CT or MR input in about 2 ms, matching the k-Wave focus location within about 3 mm.

  5. Cosmo-SPINN: Fuzzy Dark Matter Simulations with Physics-Informed Generative Networks

    astro-ph.CO 2026-07 conditional novelty 6.0 of 10

    Physics-informed generative U-Nets evolve and super-resolve fuzzy dark matter fields under Schrödinger–Poisson constraints with far less supervised data than pure data-driven baselines.

  6. When is the combined load identifiable from a stress-intensity profile? A coupled forward-inverse study on SIFBench finite-element data

    math.NA 2026-07 conditional novelty 6.0 of 10

    For fixed crack geometry, combined tension–bending–bearing load is identifiable from the SIF profile iff the three elementary load profiles are affinely independent; recovery error is controlled by the minimum singula...

  7. Diagnosing Failure Modes of Neural Operators Across Diverse PDE Families

    cs.LG 2026-01 unverdicted novelty 6.0 of 10

    A single-architecture empirical study reports degradation factors up to ~18 for neural PDE solvers under distribution shift, but its headline 'architecture-dependent failure patterns' claim is not backed by the experi...

  8. Matrix-free Neural Preconditioner for the Dirac Operator in Lattice Gauge Theory

    hep-lat 2025-09 conditional novelty 6.0 of 10

    A matrix-free neural preconditioner learns to map gauge configurations to modified configurations whose Dirac operators approximate the inverse, halving CG iterations and transferring across lattice sizes.

  9. Temperature-Aware Recurrent Neural Operator for Temperature-Dependent Anisotropic Plasticity in HCP Materials

    cond-mat.mtrl-sci 2025-08 conditional novelty 6.0 of 10

    TRNO, a time-resolution-independent recurrent neural operator, accurately surrogates temperature-dependent anisotropic plasticity of polycrystalline magnesium and enables fast multiscale simulations.

  10. Bubbleformer: Forecasting Boiling with Transformers

    cs.LG 2025-07 conditional novelty 6.0 of 10

    Bubbleformer is a spatiotemporal transformer that autonomously forecasts boiling fields across fluids and regimes, and its authors release the BubbleML 2.0 dataset to support further work.

  11. Neural Operators for Forward and Inverse Potential-Density Mappings in Classical Density Functional Theory

    physics.chem-ph 2025-06 conditional novelty 6.0 of 10

    In 1D hard-rod cDFT, Fourier neural operators learn the density-to-direct-correlation-function map more accurately than DeepONet variants and dense networks, with squared ReLU giving the best extrapolation.

  12. Phi-FEM-FNO: a new approach to train a Neural Operator as a fast PDE solver for variable geometries

    math.NA 2025-02 conditional novelty 6.0 of 10

    Combining the level-set based φ-FEM discretization with a Fourier Neural Operator yields a fast surrogate PDE solver for varying geometries, with accuracy comparable to finite elements on the tested cases.

  13. Benchmarking Multi-fidelity Neural Operators on Complex PDE Problems with Non-trivial Fidelity Differences

    cs.CE 2026-08 conditional novelty 5.0 of 10

    Across four PDE benchmarks, transfer learning from low-fidelity weights is the only multi-fidelity neural operator strategy that consistently outperforms the high-fidelity-only baseline, while direct LF-input methods ...

  14. Energy Manifold Natural Gradient Descent: Riemannian Optimization for Neural PDE Solvers

    cs.LG 2026-07 conditional novelty 5.0 of 10

    Energy Manifold Natural Gradient Descent (EMNGD) defines the energy natural gradient on a Riemannian parameter manifold and proves it equals the energy-metric projection of the function-space Newton step.

  15. DRIFT: Direct Reduced Fourier Transforms for Distributed Spectral Neural Operators

    cs.DC 2026-07 conditional novelty 5.0 of 10

    Distributed Fourier Neural Operators can compute their truncated spectra with local partial DFTs and two collectives on the kept modes, giving exact results with communication independent of grid resolution.

  16. Learning 3D Hypersonic Flow with Physics-Enhanced Neural Fields: A Case Study on the Orion Reentry Capsule

    physics.flu-dyn 2026-03 conditional novelty 5.0 of 10

    A Fourier-enhanced neural field predicts steady 3D hypersonic flow around the Orion capsule in seconds, matching CFD snapshots at held-out angles of attack better than graph networks.

  17. Automatic discovery of optimal meta-solvers for time-dependent nonlinear PDEs

    math.NA 2025-06 conditional novelty 5.0 of 10

    Automated multi-objective search over hybrid neural-operator and classical iterative solver configurations finds meta-solvers that solve benchmark time-dependent nonlinear PDEs up to 15 times faster than bare Krylov methods.

  18. Mondrian: Transformer Operators via Domain Decomposition

    cs.LG 2025-06 conditional novelty 5.0 of 10

    Mondrian applies transformer attention to subdomain-restricted functions, decoupling the model from the grid resolution.

  19. Latent Mamba Operator for Partial Differential Equations

    cs.LG 2025-05 conditional novelty 5.0 of 10

    LaMO replaces attention in latent-token neural operators with bidirectional state-space models and reports consistent accuracy gains on six PDE benchmarks.

  20. A Unified Framework for Simultaneous Parameter and Function Discovery in Differential Equations

    cs.LG 2025-05 reject novelty 4.0 of 10

    The paper proves identifiability conditions for ODE inverse problems with one unknown constant and one unknown function, and adds approximate error bounds when data points are close but not identical.

  21. Physics-Encoded Inverse Modeling for Arctic Snow Depth Estimation

    cs.LG 2026-01 reject novelty 3.0 of 10

    PhysE-Inv predicts a snow-depth proxy generated from the same ERA5 inputs the model sees, so its accuracy against real Arctic snow depth is unestablished.

  22. The Fourier Spectral Transformer Networks For Efficient and Generalizable Nonlinear PDEs Prediction

    cs.LG 2025-07 reject novelty 3.0 of 10

    The paper trains a Transformer on Fourier spectral coefficients to surrogate 1D Burgers and 2D Navier-Stokes dynamics, but its claimed superiority over numerical and ML baselines is not demonstrated.

  23. Toward Intelligent Electronic-Photonic Design Automation for Large-Scale Photonic Integrated Circuits: from Device Inverse Design to Physical Layout Generation

    cs.ET 2025-07 reject novelty 2.0 of 10

    PoLaRIS is an architectural proposal that assembles the authors' earlier BOSON-1, MAPS, Apollo and LiDAR tools into one photonic design automation pipeline, with no new results demonstrated.

Pith tools