# GCVAE: Generalized-Controllable Variational AutoEncoder

arXiv: https://arxiv.org/abs/2206.04225 · HTML: https://mlsanity.com/paper/2206.04225/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 35 references that are in the map.
- First posted on arXiv 9 Jun 2022.

## Builds on (most cited first, 13 of 13)

- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets](https://mlsanity.com/paper/1606.03657/) (2016)
- [MINE: Mutual Information Neural Estimation](https://mlsanity.com/paper/1801.04062/) (2018)
- [Disentangling by Factorising](https://mlsanity.com/paper/1802.05983/) (2018)
- [DRAW: A Recurrent Neural Network For Image Generation](https://mlsanity.com/paper/1502.04623/) (2015)
- [Disentangled Representation Learning for Non-Parallel Text Style Transfer](https://mlsanity.com/paper/1808.04339/) (2018)
- [InfoVAE: Information Maximizing Variational Autoencoders](https://mlsanity.com/paper/1706.02262/) (2017)
- [Deep Convolutional Inverse Graphics Network](https://mlsanity.com/paper/1503.03167/) (2015)
- [Learning Deep Disentangled Embeddings with the F-Statistic Loss](https://mlsanity.com/paper/1802.05312/) (2018)
- [Theory and Evaluation Metrics for Learning Disentangled Representations](https://mlsanity.com/paper/1908.09961/) (2019)
- [Deep Density-based Image Clustering](https://mlsanity.com/paper/1812.04287/) (2018)
- [Towards Generating Long and Coherent Text with Multi-Level Latent Variable Models](https://mlsanity.com/paper/1902.00154/) (2019)
- [Unsupervised Abstractive Sentence Summarization using Length Controlled Variational Autoencoder](https://mlsanity.com/paper/1809.05233/) (2018)

## Abstract (arXiv)

Variational autoencoders (VAEs) have recently been used for unsupervised disentanglement learning of complex density distributions. Numerous variants exist to encourage disentanglement in latent space while improving reconstruction. However, none have simultaneously managed the trade-off between attaining extremely low reconstruction error and a high disentanglement score. We present a generalized framework to handle this challenge under constrained optimization and demonstrate that it outperforms state-of-the-art existing models as regards disentanglement while balancing reconstruction. We introduce three controllable Lagrangian hyperparameters to control reconstruction loss, KL divergence loss and correlation measure. We prove that maximizing information in the reconstruction network is equivalent to information maximization during amortized inference under reasonable assumptions and constraint relaxation.