Learning Discrete Structured Representations by Adversarially Maximizing Mutual Information
arXiv:2004.03991
Abstract
We propose learning discrete structured representations from unlabeled data by maximizing the mutual information between a structured latent variable and a target variable. Calculating mutual information is intractable in this setting. Our key technical contribution is an adversarial objective that can be used to tractably estimate mutual information assuming only the feasibility of cross entropy calculation. We develop a concrete realization of this general formulation with Markov distributions over binary encodings. We report critical and unexpected findings on practical aspects of the objective such as the choice of variational priors. We apply our model on document hashing and show that it outperforms current best baselines based on discrete and vector quantized variational autoencoders. It also yields highly compressed interpretable representations.
ICML 2020
References in corpus (17)
- Adam: A Method for Stochastic Optimization
- Efficient Estimation of Word Representations in Vector Space
- Auto-Encoding Variational Bayes
- Representation Learning with Contrastive Predictive Coding
- Categorical Reparameterization with Gumbel-Softmax
- Neural Discrete Representation Learning
- The information bottleneck method
- Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
- Learning deep representations by mutual information estimation and maximization
- The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
- Neural Variational Inference and Learning in Belief Networks
- Fast Decoding in Sequence Models using Discrete Latent Variables
- On Variational Bounds of Mutual Information
- Variational inference for Monte Carlo objectives
- Lagging Inference Networks and Posterior Collapse in Variational Autoencoders
- Information Theoretic Co-Training
- Learning Discrete Structured Representations by Adversarially Maximizing Mutual Information