Neural Logic Reinforcement Learning

Zhengyao Jiang, Shan Luo

Deep reinforcement learning (DRL) has achieved significant breakthroughs in various tasks. However, most DRL algorithms suffer a problem of generalizing the learned policy which ma…