activity
20122024
most citedUniversal Approximation in Dropout Neural Networks

11 citations · 13 across the 8 of their papers we have counts for

collaborators

12 papers

cs.LG2024

Topological degree as a discrete diagnostic for disentanglement, with applications to the VAE

Mahefa Ratsisetraina Ravelonanosy, Vlado Menkovski, Jacobus W. Portegies

We investigate the ability of Diffusion Variational Autoencoder (VAE) with unit sphere as latent space to capture topological and geometrical structure and disen…

cs.LG2022

Neural Langevin Dynamics: towards interpretable Neural Stochastic Differential Equations

Simon M. Koop, Mark A. Peletier, Jacobus W. Portegies +1

Neural Stochastic Differential Equations (NSDE) have been trained as both Variational Autoencoders, and as GANs. However, the resulting Stochastic Differential Equations can be har…

math.DG2022

Small time asymptotics of the entropy of the heat kernel on a Riemannian manifold

Vlado Menkovski, Jacobus W. Portegies, Mahefa Ratsisetraina Ravelonanosy

We give an asymptotic expansion of the relative entropy between the heat kernel of a compact Riemannian manifold and the normalized Riemannian volume for small val…

cs.LG202011 cited

Universal Approximation in Dropout Neural Networks

Oxana A. Manita, Mark A. Peletier, Jacobus W. Portegies +2

We prove two universal approximation theorems for a range of dropout neural networks. These are feed-forward neural networks in which each edge is given a random -valued f…

cs.LG20201 cited

A Metric for Linear Symmetry-Based Disentanglement

Luis A. Pérez Rey, Loek Tonnaer, Vlado Menkovski +2

The definition of Linear Symmetry-Based Disentanglement (LSBD) proposed by (Higgins et al., 2018) outlines the properties that should characterize a disentangled representation tha…

math.PR2019

Arrow Contraction and Expansion in Tropical Diagrams

Rostislav Matveev, Jacobus W. Portegies

Arrow contraction applied to a tropical diagram of probability spaces is a modification of the diagram, replacing one of the morphisms by an isomorphims, while preserving other par…