Arxiu mensual: abril de 2018

Evolved Policy Gradients, an experimental meta-learning technique to let agents rapidly learn to solve novel tasks: https://blog.openai.com/evolved-policy-gradients/