Representation Learning in Deep RL via Discrete Information Bottleneck

Open in new window