检索结果-内蒙古大学图书馆

Exploring biologically inspired mechanisms of adversarial robustness

学校读者我要写书评

暂无评论

Neural computing and Applications 2025年 1-12页

作者： Holzhausen, Konstantin Merlid, Mia Torvik, Håkon Olav Malthe-Sørenssen, Anders Lepperød, Mikkel Elle Department of Physics University of Oslo Sem Sælands vei 24 Fysikkbygningen Oslo0371 Norway Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Kristian Augusts gate 23 Oslo0164 Norway

Backpropagation-optimized artificial neural networks, while precise, lack robustness, leading to unforeseen behaviors that affect their safety. Biological neural systems do solve some of these issues already. Unlike artificial models, biological neurons adjust connectivity based on neighboring cell activity. Understanding the biological mechanisms of robustness can pave the way toward building trustworthy and safe systems. Robustness in neural representations is hypothesized to correlate with the smoothness of the encoding manifold. Recent work suggests power law covariance spectra, which were observed studying the primary visual cortex of mice, to be indicative of a balanced trade-off between accuracy and robustness in representations. Here, we show that unsupervised local learning models with winner takes all dynamics learn such power law representations, providing upcoming studies a mechanistic model with that characteristic. Our research aims to understand the interplay between geometry, spectral properties, robustness, and expressivity in neural representations. Hence, we study the link between representation smoothness and spectrum by using weight, Jacobian and spectral regularization while assessing performance and adversarial robustness. Our work serves as a foundation for future research into the mechanisms underlying power law spectra and optimally smooth encodings in both biological and artificial systems. The insights gained may elucidate the mechanisms that realize robust neural networks in mammalian brains and inform the development of more stable and reliable artificial systems. © The Author(s) 2025.

关键词： Generative adversarial networks

Correction: Exploring biologically inspired mechanisms of adversarial robustness(Neural computing and Applications, 10.1007/s00521-025-11019-6)

学校读者我要写书评

暂无评论

Neural computing and Applications 2025年 1-2页

In this article, panel (b) of Fig. 6 was inadvertently printed in black and white instead of colour;the figure should have appeared as shown below. (Figure presented.) Norm of the Jacobian of latent activations h^ (upper panel) and decision landscape (lower panel) of studied models. Similarly to [12], the displayed planes are created by grid sampling from a random plane projection in the model’s input space The original article has been corrected. © The Author(s) 2025.

关键词：

An energy stable and conservative multiplicative dynamical low-rank discretization for the Su-Olson problem

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Baumann, Lena Einkemmer, Lukas Klingenberg, Christian Kusch, Jonas University of Wuerzburg Department of Mathematics Wuerzburg Germany University of Innsbruck Numerical Analysis and Scientific Computing Innsbruck Austria Norwegian University of Life Sciences Department of Data Science Ås Norway

computing numerical solutions of the thermal radiative transfer equations on a finely resolved grid can be costly due to high computational and memory requirements. A numerical reduced order method that has recently been applied to a wide variety of kinetic partial differential equations is the concept of dynamical low-rank approximation (DLRA). In this paper, we consider the thermal radiative transfer equations with Su-Olson closure, leading to a linearized kinetic model. For the conducted theoretical and practical considerations we use a multiplicative splitting of the distribution function that poses additional challenges in finding an energy stable discretization and deriving a hyperbolic Courant-Friedrichs-Lewy (CFL) condition. We propose such an energy stable DLRA scheme that makes use of the augmented basis update & Galerkin integrator. This integrator allows for additional basis augmentations, enabling us to give a mathematically rigorous proof of energy stability and local mass conservation. numerical examples confirm the derived properties and show the computational advantages of the DLRA scheme compared to a numerical solution of the full system of *** Codes 35L65, 35Q49, 65M12, 65M22 © 2025, CC BY-NC-ND.

关键词： Radiative transfer

On the numerical evaluation of wall shear stress using the finite element method

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Brunátová, Jana Dokken, Jørgen Schartum Valen-Sendstad, Kristian Hron, Jaroslav Mathematical Institute Charles University Czech Republic Bernoulli Institute University of Groningen Netherlands Scientific Computing and Numerical Analysis Simula Research Laboratory Norway Computational Physiology Simula Research Laboratory Norway

Wall shear stress (WSS) is a crucial hemodynamic quantity extensively studied in cardiovascular research, yet its numerical computation is not straightforward. This work aims to compare WSS results obtained from two different finite element discretizations, quantify the differences between continuous and discontinuous stresses, and introduce a novel method for WSS evaluation through the formulation of a boundary-flux problem. Two benchmark problems are considered – a 2D Stokes flow on a unit square and a 3D Poiseuille flow through a cylindrical pipe. These are followed by investigations of steady-state Navier–Stokes flow in two patient-specific aneurysms. The study focuses on P1/P1 stabilized and Taylor–Hood P2/P1 mixed finite elements for velocity and pressure. WSS is computed using either the proposed boundary-flux method or as a projection of tangential traction onto First order Lagrange (P1), Discontinuous Galerkin first order (DG-1), or Discontinuous Galerkin zero order (DG-0) space. For the P1/P1 stabilized element, the boundary-flux and P1 projection methods yielded equivalent results. With the P2/P1 element, the boundary-flux evaluation demonstrated faster convergence in the Poiseuille flow example but showed increased sensitivity to pressure field inaccuracies in patient-specific geometries compared to the projection method. In patient-specific cases, the P2/P1 element exhibited superior robustness to mesh size when evaluating average WSS and low shear area (LSA), outperforming the P1/P1 stabilized element. Projecting discontinuous finite element results into continuous spaces can introduce artifacts, such as the Gibbs phenomenon. Consequently, it becomes crucial to carefully select the finite element space for boundary stress calculations – not only in applications involving WSS computations for aneurysms. © 2025, CC BY.

关键词： Shear stress

Hexagons all the way down: grid cells as a conformal isometric map of space

学校读者我要写书评

暂无评论

PLoS Computational Biology 2025年第2期21卷 e1012804页

作者： Schøyen, Vemund Sigmundson Beshkov, Kosio Pettersen, Markus Borud Hermansen, Erik Holzhausen, Konstantin Malthe-Sørenssen, Anders Fyhn, Marianne Lepperød, Mikkel Elle Department of Biosciences University of Oslo Oslo Norway Department of Physics University of Oslo Oslo Norway Department of Mathematical Sciences Norwegian University of Science and Technology Trondheim Norway Department of Numerical Analysis and Scientific Computing Simula Research Laboratory AS Oslo Norway

Grid cells in the entorhinal cortex are known for their hexagonal spatial activity patterns and are thought to provide a neural metric for space, and support path integration. In this study, we further investigate grid cells as a metric of space by optimising them for a conformal isometric (CI) map of space using a model based on a superposition of plane waves. By optimising the phases within a single grid cell module, we find that the module can form a CI of two-dimensional flat space with phases arranging into a regular hexagonal pattern, supporting an accurate spatial metric. Additionally, we find that experimentally recorded grid cells exhibit CI properties, with one example module showing a phase arrangement similar to the hexagonal pattern observed in our model. These findings provide computational and preliminary experimental support for grid cells as a CI-based spatial representation. We also examine other properties that emerge in CI-optimised modules, including consistent energy expenditure across space and the minimal cell count required to support unique representation of space and maximally topologically persistent toroidal population activity. Altogether, our results suggest that grid cells are well-suited to form a CI map, with several beneficial properties arising from this organisation. © 2025 Schøyen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

关键词： Cytology

FULLY DISCRETE analysis OF THE GALERKIN POD NEURAL NETWORK APPROXIMATION WITH APPLICATION TO 3D ACOUSTIC WAVE SCATTERING

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Dölz, Jürgen Henríquez, Fernando Institute for Numerical Simulation University of Bonn Friedrich-Hirzebruch-Allee 7 Bonn53115 Germany Institute for Analysis and Scientific Computing Vienna University of Technology Wiedner Hauptstrae 8-10 WienA-1040 Austria

In this work, we consider the approximation of parametric maps using the so-called Galerkin POD-NN method. This technique combines the computation of a reduced basis via proper orthogonal decomposition (POD) and artificial neural networks (NNs) for the construction of fast surrogates of said parametric maps. We provide a fully discrete error analysis of this method that accounts for different discretization errors, including the number of reduced basis in the approximation of the solution manifold, truncation in the parameter space, and, most importantly, the number of samples in the computation of the reduced space, together with the effect of the use of NNs in the approximation of the reduced coefficients. Following this error analysis we provide a-priori bounds on the required POD tolerance, the resulting POD ranks, and NN-parameters to maintain the order of convergence of quasi Monte Carlo sampling techniques. The main advantage of Galerkin POD-NN over existing traditional techniques is that the online and offline phases of the projection-based reduced basis method are completely decoupled, making the evaluation of the constructed surrogate in run-time considerably faster without the need for an hyper-reduction technique. We conclude this work by showcasing the applicability of this method through a practical industrial application: the sound-soft acoustic scattering problem by a parametrically defined scatterer in three physical dimensions. Copyright © 2025, The Authors. All rights reserved.

关键词： Monte Carlo methods

PARAMETER-ROBUST PRECONDITIONERS FOR A FOUR-FIELD THERMO-POROELASTICITY MODEL

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Cai, Mingchao Kuchta, Miroslav Li, Jingzhi Li, Ziliang Mardal, Kent-Andre Department of Mathematics Morgan State University BaltimoreMD21251 United States Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Oslo0164 Norway Department of Mathematics Southern University of Science and Technology Shenzhen518055 China

We study a thermo-poroelasticity model which describes the interaction between the deformation of an elastic porous material and fluid flow under non-isothermal conditions. The model involves several parameters that can vary significantly in practical applications, posing a challenge for developing discretization techniques and solution algorithms that handle such variations effectively. We propose a four-field formulation and apply a conforming finite element discretization. The primary focus is on constructing and analyzing preconditioners for the resulting linear system. Two preconditioners are proposed: one involves regrouping variables and treating the 4-by-4 system as a 2-by-2 block form, while the other is directly constructed from the 4-by-4 coupled operator. Both preconditioners are demonstrated to be robust with respect to variations in parameters and mesh refinement. numerical experiments are presented to demonstrate the effectiveness of the proposed preconditioners and validate their theoretical performance under varying parameter *** Codes 65M60, 65F08, 65F10 Copyright © 2025, The Authors. All rights reserved.

关键词： Flow of fluids

THE LATENT VARIABLE PROXIMAL POINT ALGORITHM FOR VARIATIONAL PROBLEMS WITH INEQUALITY CONSTRAINTS

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Dokken, Jørgen S. Farrell, Patrick E. Keith, Brendan Papadopoulos, Ioannis P.A. Surowiec, Thomas M. Department of Scientific Computing and Numerical Analysis Simula Research Laboratory Oslo Norway Mathematical Institute University of Oxford Oxford United Kingdom Mathematical Institute Faculty of Mathematics and Physics Charles University Prague Czech Republic Division of Applied Mathematics Brown University ProvidenceRI United States Weierstrass Institute Berlin Germany

The latent variable proximal point (LVPP) algorithm is a framework for solving infinite-dimensional variational problems with pointwise inequality constraints. The algorithm is a saddle point reformulation of the Bregman proximal point algorithm. At the continuous level, the two formulations are equivalent, but the saddle point formulation is more amenable to discretization because it introduces a structure-preserving transformation between a latent function space and the feasible set. Working in this latent space is much more convenient for enforcing inequality constraints than the feasible set, as discretizations can employ general linear combinations of suitable basis functions, and nonlinear solvers can involve general additive updates. LVPP yields numerical methods with observed mesh-independence for obstacle problems, contact, fracture, plasticity, and others besides;in many cases, for the first time. The framework also extends to more complex constraints, providing means to enforce convexity in the Monge–Ampère equation and handling quasi-variational inequalities, where the underlying constraint depends implicitly on the unknown solution. In this paper, we describe the LVPP algorithm in a general form and apply it to twelve problems from across mathematics. © 2025, CC BY.

关键词： Consensus algorithm

PARAMETER-ROBUST PRECONDITIONER FOR STOKES-DARCY COUPLED PROBLEM WITH LAGRANGE MULTIPLIER

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hu, Xiaozhe Kuchta, Miroslav Mardal, Kent-Andre Wang, Xue Department of Mathematics Tufts University MedfordMA02155 United States Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Oslo0164 Norway Simula Research Laboratory Oslo0164 Norway Department of Mathematics University of Oslo Blindern Oslo0316 Norway School of Mathematics Shandong University Jinan250100 China

In this paper, we propose a parameter-robust preconditioner for the coupled Stokes-Darcy problem equipped with various boundary conditions, enforcing the mass conservation at the interface via a Lagrange multiplier. We rigorously establish that the coupled system is well-posed with respect to physical parameters and mesh size and provide a framework for constructing parameter-robust preconditioners. Furthermore, we analyze the convergence behavior of the Minimal Residual method in the presence of small outlier eigenvalues linked to specific boundary conditions, which can lead to slow convergence or stagnation. To address this issue, we employ deflation techniques to accelerate the convergence. Finally, numerical experiments confirm the effectiveness and robustness of the proposed approach. Copyright © 2025, The Authors. All rights reserved.

关键词： Lagrange multipliers