Bridging Spectral Operator Learning and U-Net Hierarchies: SpectraNet for Stable Autoregressive PDE Surrogates
Not on the map. Why?
Abstract (arXiv)
Neural operators for time-dependent PDEs face a structural tension: spectral architectures (FNO and descendants) inherit exponential rollout-error growth from their one-step Lipschitz constant, while hierarchical U-Net operators trade resolution invariance for multi-scale detail. We introduce SpectraNet, an autoregressive neural operator that composes truncated spectral convolutions inside a U-Net hierarchy with a Residual-Target Spectral Block trained under a Semigroup-Consistency Loss. The residual-target parametrization replaces L^T stability blow-up with linear T*delta drift, and the spectral path's parameter count is Theta(L w^2 M^2), independent of grid N. Under a single unified protocol against 16 published neural-operator baselines on Navier-Stokes nu=1e-5 at 64x64, SpectraNet reaches test relative L2 = 0.0822 at 2.04M parameters -- 2.33x fewer than canonical FNO at ~20% lower error -- and wins five of six rows in a cross-PDE comparison against FNO (NS at nu in {1e-4, 1e-3}, PDEBench Shallow-Water 2D and Diffusion-Reaction, with the Active-Matter row going to FNO inside its seed spread). Trained from scratch at native 128^2 under the same protocol, SpectraNet improves to 0.0724 while FNO regresses to 0.3080. Free rollout stays bounded for T=100 where FNO diverges across all 200 test trajectories. On consumer CPU at B=1, SpectraNet runs sub-200ms while the full-attention Transformer that wins raw L2 pays ~60x latency; we do not claim to beat that Transformer on raw L2, only to dominate the lightweight (<=5M parameter, sub-200ms CPU) Pareto frontier. Source code: https://github.com/Enrikkk/spectranet
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 21 of its 29 references that are in the map. First posted on arXiv 9 May 2026.
Builds on
21 of 29 references.
References · 21
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
- 2020Fourier Neural Operator for Parametric Partial Differential Equationscited by 2,908
- 2024KAN: Kolmogorov-Arnold Networkscited by 649
- 2021Learning the solution operator of parametric partial differential equations with physics-informed DeepOnetscited by 540
- 2021Neural Operator: Learning Maps Between Function Spacescited by 414
- 2022PDEBENCH: An Extensive Benchmark for Scientific Machine Learningcited by 413
- 2021Choose a Transformer: Fourier or Galerkincited by 343
- 2023GNOT: A General Neural Operator Transformer for Operator Learningcited by 336
- 2024Transolver: A Fast Transformer Solver for PDEs on General Geometriescited by 326
- 2021U-FNO -- An enhanced Fourier neural operator-based deep-learning model for multiphase flowcited by 297
- 2022Transformer for Partial Differential Equations' Operator Learningcited by 283
9 more
- 2021Factorized Fourier Neural Operatorscited by 249
- 2021Multiwavelet-based Operator Learning for Differential Equationscited by 234
- 2022U-NO: U-shaped Neural Operatorscited by 230
- 2023PDE-Refiner: Achieving Accurate Long Rollouts with Neural PDE Solverscited by 216
- 2023Convolutional Neural Operators for robust and accurate learning of PDEscited by 214
- 2024The Well: a Large-Scale Collection of Diverse Physics Simulations for Machine Learningcited by 145
- 2023Scalable Transformer for PDE Surrogate Modelingcited by 128
- 2023Solving High-Dimensional PDEs with Latent Spectral Modelscited by 89
- 2023Improved Operator Learning by Orthogonal Attentioncited by 61
Built on by
Citing papers per year: 2026: 1.
1 paper.