mlsanityData to 14 Oct 2026

Controlling Posterior Collapse by an Inverse Lipschitz Constraint on the Decoder Network

arXiv 2304.12770 ↗First posted 25 Apr 2023cs.LG stat.MLcited by 7

Not on the map. Why?

Abstract (arXiv)

Variational autoencoders (VAEs) are one of the deep generative models that have experienced enormous success over the past decades. However, in practice, they suffer from a problem called posterior collapse, which occurs when the encoder coincides, or collapses, with the prior taking no information from the latent structure of the input data into consideration. In this work, we introduce an inverse Lipschitz neural network into the decoder and, based on this architecture, provide a new method that can control in a simple and clear manner the degree of posterior collapse for a wide range of VAE models equipped with a concrete theoretical guarantee. We also illustrate the effectiveness of our method through several numerical experiments.

Key facts. Cited by 7 papers in the map (data through 14 Oct 2026). Builds on 33 of its 55 references that are in the map. First posted on arXiv 25 Apr 2023.

Builds on

33 of 55 references.

References · 33

  1. 2017Neural Discrete Representation Learningcited by 4,712
  2. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,349
  3. 2015Generating Sentences from a Continuous Spacecited by 612
  4. 2015Importance Weighted Autoencoderscited by 456
  5. 2016Input Convex Neural Networkscited by 414
  6. 2016Ladder Variational Autoencoderscited by 318
  7. 2017VAE with a VampPriorcited by 248
  8. 2019Rapid Convergence of the Unadjusted Langevin Algorithm: Isoperimetry Sufficescited by 227
  9. 2016Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoderscited by 220
  10. 2019Cyclical Annealing Schedule: A Simple Approach to Mitigating KL Vanishingcited by 191
  11. 2016Variational Lossy Autoencodercited by 176
  12. 2019Lagging Inference Networks and Posterior Collapse in Variational Autoencoderscited by 124
21 more
  1. 2019Don't Blame the ELBO! A Linear VAE Perspective on Posterior Collapsecited by 121
  2. 2019Preventing Posterior Collapse with delta-VAEscited by 86
  3. 2018Optimal Control Via Neural Networks: A Convex Approachcited by 85
  4. 2018Semi-Amortized Variational Autoencoderscited by 83
  5. 2017Improved Variational Autoencoders for Text Modeling using Dilated Convolutionscited by 81
  6. 2016PixelVAE: A Latent Variable Model for Natural Imagescited by 70
  7. 2023Posterior Collapse and Latent Variable Non-identifiabilitycited by 57
  8. 2019The Usual Suspects? Reassessing Blame for VAE Posterior Collapsecited by 52
  9. 2018Avoiding Latent Variable Collapse With Generative Skip Modelscited by 47
  10. 2017Assigning a value to a power likelihood in a general Bayesian modelcited by 46
  11. 2018Unsupervised Discrete Sentence Representation Learning for Interpretable Neural Dialog Generationcited by 34
  12. 2019A Surprisingly Effective Fix for Deep Latent Variable Modeling of Textcited by 32
  13. 2018Improving Explorability in Variational Inference with Annealed Variational Objectivescited by 24
  14. 2022Posterior Collapse of a Linear Latent Variable Modelcited by 18
  15. 2017Tackling Over-pruning in Variational Autoencoderscited by 14
  16. 2021Certifiably Robust Variational Autoencoderscited by 14
  17. 2019Variational approximations using Fisher divergencecited by 13
  18. 2018Practical bounds on the error of Bayesian posterior approximations: A nonasymptotic approachcited by 12
  19. 2020Preventing Posterior Collapse with Levenshtein Variational Autoencodercited by 12
  20. 2020Fisher Auto-Encoderscited by 5
  21. 2020Improving k-Means Clustering Performance with Disentangled Internal Representationscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 0; 2025: 3; 2026: 3.

7 papers.

  1. 2023Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoderscited by 7
  2. 2025A Controllable Appearance Representation for Flexible Transfer and Editingcited by 2
  3. 2025S2WTM: Spherical Sliced-Wasserstein Autoencoder for Topic Modelingcited by 1
  4. 2025Toward Architecture-Agnostic Local Control of Posterior Collapse in VAEscited by 0
  5. 2026Quasi Monte Carlo methods enable extremely low-dimensional deep generative modelscited by 0
  6. 2026Robust Provably Secure Image Steganography via Latent Iterative Optimizationcited by 0
  7. 2026RST: certifying and constructing prescribed information in variational autoencoderscited by 0