StarCraft II is one of the most challenging simulated reinforcement lear...
We introduce DeepNash, an autonomous agent capable of learning to play t...
One of the key promises of model-based reinforcement learning is the abi...
Data-efficiency and generalization are key challenges in deep learning a...
Recent developments in the field of model-based RL have proven successfu...
Deep networks have achieved excellent results in perceptual tasks, yet t...
State representation learning, or the ability to capture latent generati...
Unsupervised exploration and representation learning become increasingly...
Estimating and optimizing Mutual Information (MI) is core to many proble...
Mutual information maximization has emerged as a powerful learning objec...
We show that an end-to-end deep learning approach can be used to recogni...
For discrete data, the likelihood P(x) can be rewritten exactly and
para...
We propose a new framework for estimating generative models via an
adver...
Generative Stochastic Networks (GSNs) have been recently introduced as a...