Overcoming Spectral Bias via Cross-Attention
math.NA, cs.NA
Submitted: 2025-12-21
Updated: 2026-11-05
Comments: 30 pages, 21 figures
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Frequency Principle: Fourier Analysis Sheds Light on Deep Neural Networks
- Fourier Multi-Component and Multi-Layer Neural Networks: Unlocking High-Frequency Potential
- Why Shallow Networks Struggle to Approximate and Learn High Frequencies
- On the Activation Function Dependence of the Spectral Bias of Neural Networks
- The Random Feature Method for Time-dependent Problems
- Multi-scale Deep Neural Network (MscaleDNN) for Solving Poisson-Boltzmann Equation in Complex Domains
- Multi-scale Deep Neural Network (MscaleDNN) Methods for Oscillatory Stokes Flows in Complex Domains
- Generative AI for fast and accurate statistical computation of fluids
- Adaptive Fourier Neural Operators: Efficient Token Mixers for Transformers
- FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
- Frequency-adaptive tensor neural networks for high-dimensional multi-scale problems
- Data-integrated neural networks for solving partial differential equations
Related papers
- Do physics-informed neural networks (PINNs) need to be deep? Shallow PINNs using the Levenberg-Marquardt algorithm
- A Neural-preconditioned Poisson Solver for Mixed Dirichlet and Neumann Boundary Conditions
- Advanced Linear Algebra with Applications - Part I (Numerical linear algebra for PDEs, machine learning, and data assimilation)
- Operator learning meets inverse problems: A probabilistic perspective
- Second-order consistency for learning chaotic dynamics via randomized Jacobian matching
- Windowed thinning and query complexity for the bouncy particle and Zigzag samplers