paper-with-me

홈 › Papers

Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias

2021-12-13 · ICLR 2022 4 · Frederic Koehler, Viraj Mehta, Chenghui Zhou, Andrej Risteski

Variational Autoencoders are one of the most commonly used generative models, particularly for image data. A prominent difficulty in training VAEs is data that is supported on a lower-dimensional manifold. Recent work by Dai and Wipf (2020) proposes a two-stage training algorithm for VAEs, based on a conjecture that in standard VAE training the generator will converge to a solution with 0 variance which is correctly supported on the ground truth manifold. They gave partial support for that conjecture by showing that some optima of the VAE loss do satisfy this property, but did not analyze the training dynamics. In this paper, we show that for linear encoders/decoders, the conjecture is true-that is the VAE training does recover a generator with support equal to the ground truth manifold-and does so due to an implicit bias of gradient descent rather than merely the VAE loss itself. In the nonlinear case, we show that VAE training frequently learns a higher-dimensional manifold which is a superset of the ground truth manifold.

📄 PDF Abstract BibTeX arXiv:2112.06868

Code (1)

virajmehta/vae-training 공식 구현 jax

Similar Papers 제목 키워드 기반

Interpretable Embeddings From Molecular Simulations Using Gaussian Mixture Variational Autoencoders

2019-12-22 · Yasemin Bozkurt Varolgunes, Tristan Bereau, Joseph F. Rudzinski

Extracting insight from the enormous quantity of data generated from molecular simulations requires the identification of a small number of collective variables whose corresponding low-dimensional free-energy landscape r…

ClusteringDimensionality ReductionPhysical Intuition

Variational Composite Autoencoders

2018-04-12 · Jiangchao Yao, Ivor Tsang, Ya zhang

Learning in the latent variable model is challenging in the presence of the complex data structure or the intractable latent variable. Previous variational autoencoders can be low effective due to the straightforward enc…

Decoder

Perturbation theory approach to study the latent space degeneracy of Variational Autoencoders

2019-07-10 · Helena Andrés-Terré, Pietro Lió

The use of Variational Autoencoders in different Machine Learning tasks has drastically increased in the last years. They have been developed as denoising, clustering and generative tools, highlighting a large potential …

BIG-bench Machine LearningClusteringDenoising

Variational embedding of protein folding simulations using gaussian mixture variational autoencoders

2021-08-27 · Mahdi Ghorbani, Samarjeet Prasad, Jeffery B. Klauda, Bernard R. Brooks

Conformational sampling of biomolecules using molecular dynamics simulations often produces large amount of high dimensional data that makes it difficult to interpret using conventional analysis techniques. Dimensionalit…

Dimensionality ReductionProtein Folding

FONDUE: an algorithm to find the optimal dimensionality of the latent representations of variational autoencoders

2022-09-26 · Lisa Bonheme, Marek Grzes

When training a variational autoencoder (VAE) on a given dataset, determining the optimal number of latent variables is mostly done by grid search: a costly process in terms of computational time and carbon footprint. In…