A collection of implementations of various agents used in the paper Thinking Fast and Slow with Deep Learning and Tree Search . Currently features the following agents:
All agents run in the minihex environment.