Pith. sign in

REVIEW 9 cited by

Neural Operators with Localized Integral and Differential Kernels

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 2402.16845 v2 pith:YTMCDRCZ submitted 2024-02-26 cs.LG cs.AIcs.NAmath.NA

Miguel Liu-Schiaffini , Julius Berner , Boris Bonev , Thorsten Kurth , Kamyar Azizzadenesheli , Anima Anandkumar This is my paper · ORCID
classification cs.LGcs.AIcs.NAmath.NA
keywords operatorslearninglocalneuralcapturedifferentialintegralkernels
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Neural operators learn mappings between function spaces, which is practical for learning solution operators of PDEs and other scientific modeling applications. Among them, the Fourier neural operator (FNO) is a popular architecture that performs global convolutions in the Fourier space. However, such global operations are often prone to over-smoothing and may fail to capture local details. In contrast, convolutional neural networks (CNN) can capture local features but are limited to training and inference at a single resolution. In this work, we present a principled approach to operator learning that can capture local features under two frameworks by learning differential operators and integral operators with locally supported kernels. Specifically, inspired by stencil methods, we prove that we obtain differential operators under an appropriate scaling of the kernel values of CNNs. To obtain local integral operators, we utilize suitable basis representations for the kernels based on discrete-continuous convolutions. Both these approaches preserve the properties of operator learning and, hence, the ability to predict at any resolution. Adding our layers to FNOs significantly improves their performance, reducing the relative L2-error by 34-72% in our experiments, which include a turbulent 2D Navier-Stokes and the spherical shallow water equations.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. FourCastNet 3: A geometric approach to probabilistic machine-learning weather forecasting at scale

    cs.LG 2025-07 conditional novelty 7.0 of 10

    A purely convolutional, spherical-geometry weather model trained with a combined spatial and spectral CRPS loss delivers GenCast-level skill, IFS-beating accuracy, and stable spectra out to 60 days.

  2. Autoregressive One-Step Generative Modeling for Dynamical System Forecasting

    cs.LG 2026-05 unverdicted novelty 6.0 of 10

    MeLISA extends pixel-space MeanFlow to one-step window-conditioned autoregressive forecasting, improving long-horizon turbulence statistics over neural-operator baselines.

  3. Adaptive Physics Transformer with Fused Global-Local Attention for Subsurface Energy Systems

    cs.LG 2026-02 conditional novelty 6.0 of 10

    APT, a mesh-agnostic neural operator fusing graph-based local features with global attention, is claimed to be the first architecture trained directly on adaptive-mesh-refinement simulations and outperforms state-of-t...

  4. Modeling turbulent and self-gravitating fluids with Fourier neural operators

    astro-ph.GA 2025-07 conditional novelty 6.0 of 10

    Fourier neural operators trained on 2D projected views of 3D astrophysical simulations can forecast subsequent projected density and velocity snapshots with 5-25% RMS error, though a constant hidden magnetic field mea...

  5. A Learning-based Domain Decomposition Method

    cs.LG 2025-07 conditional novelty 6.0 of 10

    A single pretrained neural operator can act as a reusable local solver inside additive Schwarz iterations, solving elliptic PDEs with random microstructures on large, non-convex domains.

  6. Diffeomorphic Neural Operator Learning

    math.NA 2025-08 unverdicted novelty 5.0 of 10

    A neural operator that evolves fields by composing learned diffeomorphisms, enforcing relabeling symmetry and targeting conservative, non-diffusive turbulent forecasts.

  7. Accelerating PDE-Constrained Optimization by the Derivative of Neural Operators

    cs.LG 2025-06 conditional novelty 5.0 of 10

    Reference neural operators with a Virtual-Fourier layer learn solution derivatives and a hybrid solver feedback loop accelerates PDE-constrained optimization.

  8. Principled Approaches for Extending Neural Architectures to Function Spaces for Operator Learning

    cs.LG 2025-06 conditional novelty 4.0 of 10

    A practical recipe to convert common neural architectures into discretization-agnostic neural operators, validated by Navier-Stokes experiments showing cross-resolution generalization of FNO-style models.

  9. Machine learning for modelling unstructured grid data in computational physics: a review

    cs.LG 2025-02 conditional novelty 2.0 of 10

    A broad review of machine learning techniques for modeling unstructured mesh data in computational physics, with a taxonomy, a qualitative comparison, and a list of public benchmarks.

Pith tools