Deeponet
PulseAugur coverage of Deeponet — every cluster mentioning Deeponet across labs, papers, and developer communities, ranked by signal.
- 2026-08-20 research_milestone A new surrogate model, the Rationally Enriched Chebyshev (REC) trunk for DeepONets, has been developed to improve predictions for high-Péclet transport problems. source
3 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…
-
Transformer DeepONets advance radiative transfer equation solving
Researchers have developed a novel transformer-based DeepONet architecture designed to improve the accuracy and flexibility of solving the radiative transfer equation (RTE), a critical component in fire simulations. Thi…
-
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 NAViLoss objective boosts AUV velocity estimation accuracy by 44%
Researchers have developed NAViLoss, a novel objective function designed to improve the accuracy of velocity estimation for autonomous underwater vehicles (AUVs). This new method addresses limitations in existing learni…
-
Energy efficiency in neural operators for virtual sensing explored
Researchers have explored methods to reduce the energy consumption of neural operators used for virtual sensing, a process that reconstructs physical fields from observations. Their study found that techniques like comp…
-
Deep learning framework enables cross-physics mapping between distinct physical domains
Researchers have introduced Cross-Physics Mapping (CPM), a novel framework for operator learning that enables deep learning models to translate between physical domains governed by different equations. The study propose…
-
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…
-
New PI-CP method enhances uncertainty quantification for neural operators
Researchers have developed a new method called Physics-Informed Conformal Prediction (PI-CP) to provide reliable uncertainty estimates for neural operators used in approximating solutions to partial differential equatio…
-
DeepONets: Attention Mechanisms Crucial for PDE Solving Accuracy
Researchers have conducted a controlled study on Deep Neural Operators (DeepONets) to understand the impact of various attention mechanisms on their performance. The study systematically evaluated five DeepONet variants…
-
New benchmark tests operator learning for complex fluid dynamics
Researchers have developed a new benchmark for operator learning models in fluid dynamics, specifically focusing on three-phase interfacial flow. This benchmark utilizes a ternary Cahn-Hilliard-Navier-Stokes solver to g…
-
AI models advance high-resolution precipitation nowcasting
Two new research papers propose advanced deep learning architectures for high-resolution precipitation nowcasting. The first, GenONet, utilizes a Generative Adversarial Network (GAN) framework combined with a Deep Opera…
-
New Hypergraph Operator Boosts AI Accuracy for Scientific Simulations
Researchers have developed a novel method called the Hypergraph Adaptive Wavelet Operator (HALO) designed to improve the accuracy and stability of neural operators for scientific simulations. HALO operates on hypergraph…
-
Physics Attention Transformer accelerates fusion instability prediction
Researchers have developed a Physics Attention Transformer (PAT) model to rapidly predict the growth rate of vertical instabilities in fusion reactors like Alcator C-Mod and SPARC. This transformer-based approach signif…
-
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 DeepONet Surrogate Model Enhances Transport Problem Predictions
Researchers have developed a new surrogate model called the Rationally Enriched Chebyshev (REC) trunk for DeepONets, designed to handle high-Péclet transport problems with thin boundary layers. This REC trunk integrates…
-
New FAST-DeepONet method enhances AI stability for complex PDE problems
Researchers have developed FAST-DeepONet, a novel approach to improve the statistical stability of Deep Operator Networks when dealing with high-dimensional inputs from partial differential equations (PDEs). This new me…
-
New PIKFNO framework enhances neural operator interpretability
Researchers have introduced the Physics Informed Kernel Function Neural Operator (PIKFNO), a novel framework designed to enhance the interpretability of neural operators. Unlike existing methods like DeepONet that impli…
-
New Transformer Architecture Enhances Operator Learning on Complex Geometries
Researchers have introduced ArGEnT, a novel geometry-encoded Transformer designed for operator learning on arbitrary geometries. This framework decouples geometry encoding from query-point evaluation, enabling mesh-inde…
-
New MH-RG DeepONet architecture improves nonlinear wave dynamics modeling
Researchers have developed a novel Multi-Head Residual-Gated DeepONet (MH-RG) architecture designed to better model coherent nonlinear wave dynamics. This new framework integrates compact physical descriptors of the ini…
-
New research explores uncertainty quantification in deep learning for diverse applications
Three new research papers explore advanced techniques for uncertainty quantification in deep learning models. The first paper introduces intuitionistic fuzzy deep randomized neural networks (IF-dRVFL and IF-edRVFL) to i…