Fourier Neural Operators
PulseAugur coverage of Fourier Neural Operators — every cluster mentioning Fourier Neural Operators across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
New neural operator combines DMD and deep learning for efficient PDE modeling
Researchers have developed a novel neural operator that integrates dynamic mode decomposition (DMD) with deep learning to efficiently model spatiotemporal processes. This approach aims to reduce the significant computat…
-
New benchmark reveals significant error amplification in PDE foundation models
A new research paper published on arXiv introduces a benchmark for evaluating the rollout reliability of foundation models used in solving Partial Differential Equations (PDEs). The study highlights that while these mod…
-
New Bregman viewpoint advances neural operator research
Researchers have introduced a new perspective on neural operators by framing their layer actions as solutions to Bregman regularized optimization problems within Banach function spaces. This approach allows for the inte…
-
Operator Surrogates: A New Approach to Approximating Complex Operators
This paper introduces operator surrogates, a method for creating finite-parametric maps that approximate complex operators between infinite-dimensional function spaces. It details the encoder-approximator-decoder paradi…
-
New hybrid neural network models ultrasonic wave propagation
Researchers have developed a novel neural surrogate model for ultrasonic wave propagation in heterogeneous media, addressing limitations of existing methods. The new model integrates Fourier Neural Operators (FNOs) with…
-
New paper proposes physics sandbox for AI scaling limits
A new paper proposes a theoretical framework for analyzing AI scaling limits by modeling spacetime as a discrete geometric physics sandbox. This approach aims to inject non-ergodic insight into AI development, moving be…
-
New diagnostic tool OperatorCLIP probes text understanding in neural PDE models
Researchers have developed OperatorCLIP to investigate whether text conditioning in neural PDE surrogates genuinely reflects semantic understanding. Experiments using Darcy2D, ShallowWater2D, and CNS3D models showed tha…
-
New research enhances neural PDE solvers with adaptive representations and transfer learning · 6 sources tracked
Recent research explores novel methods for improving neural network solvers of partial differential equations (PDEs). One approach, "Physics-Informed Neural Plasticity," introduces a paradigm where the neural network's …
-
LSR-Net architecture learns nonlinear fluid dynamics evolution
Researchers have introduced LSR-Net, a novel neural operator architecture designed to model the forward evolution of nonlinear fluid dynamics. This network effectively learns the evolution operator from initial and futu…
-
Fourier Neural Operators Learn Lyapunov Functions for Nonlinear Systems
Researchers have developed a method to approximate Lyapunov functions for nonlinear dynamical systems using Fourier Neural Operators (FNOs). This approach aims to overcome the challenge of finding Lyapunov functions, wh…
-
New Transformer-Enhanced Neural Operator Predicts Aerodynamic Performance
Researchers have developed a new method for predicting the aerodynamic performance of turbomachinery cascades, which functions similarly to a CFD simulator. This framework first predicts fundamental parameters of the Na…
-
Wavelet-Encoded FNOs Accelerate Metamaterial Design Simulations
Researchers have developed a novel approach using Wavelet-Encoded Fourier Neural Operators (FNOs) to solve complex eigenvalue problems in physics, specifically for metamaterial design. This method effectively predicts m…
-
Neural Operators accelerate FitzHugh-Nagumo dynamics modeling
Researchers have developed parameter-conditioned Fourier Neural Operators (FNOs) to create fast, differentiable surrogate models for the FitzHugh-Nagumo (FHN) system. These models can accurately simulate neuronal voltag…
-
New research explores adaptable and domain-independent neural operators · 4 sources tracked
Researchers are exploring new methods for neural operators, which are used to approximate physical simulations. One approach, LatentDDM, focuses on pretraining operators on smaller subdomains and then using a lightweigh…
-
New MPNO model enhances stability in transient dynamics prediction
Researchers have developed a new neural operator called the Constitutive Markov Physics-Informed Neural Operator (MPNO) designed to improve stability in predicting transient dynamics, particularly for problems with stro…
-
AI models benchmarked for real-time tokamak plasma prediction
Researchers have developed an AI surrogate modeling framework to predict tokamak plasma equilibrium in real-time, addressing the computational cost of traditional Grad-Shafranov solvers. The study benchmarks five neural…
-
New 'wrong-physics backdoors' found in neural PDE operators
Researchers have identified a new vulnerability in neural partial differential equation (PDE) operators, termed "wrong-physics backdoors." This attack exploits reusable solver archives by subtly altering inputs to trigg…
-
New S2RL framework mitigates spectral bias in neural operators for underwater acoustics
Researchers have developed a new framework called Spectral-Spatial Residual Learning (S2RL) to address the spectral bias issue in Fourier Neural Operators (FNOs). This bias causes FNOs to produce over-smoothed predictio…
-
New research tackles causal inference challenges in networked systems
Two new research papers explore causal inference in complex systems where effects can spread or interfere. The first paper, "Causal Inference under Interference with Learned Exposure Mappings," investigates how uncertai…
-
New pipeline generates ML-ready datasets for 3D fluid dynamics simulations
Researchers have developed ChannelFlow-Tools, an open-source pipeline designed to generate machine-learning-ready datasets for 3D obstructed channel flows. This configuration-driven system integrates procedural obstacle…