AlphaGo

Intro vid

Combine deep neural networks with Reinforcement learning

DQN source code (paper). See Value function approximation

A3C

https://github.com/dennybritz/reinforcement-learning

https://syncedreview.com/2017/02/24/david-silver-google-deepmind-deep-reinforcement-learning/

Deep symbolic reinforcement learning combines it with symbolic reasoning.

Intentional-Unintentional agent

https://arxiv.org/abs/1611.05397