Open-Ended Reinforcement Learning with Neural Reward Functions

Robert Meier, Asier Mujika

Inspired by the great success of unsupervised learning in Computer Vision and Natural Language Processing, the Reinforcement Learning community has recently started to focus more o…