README.md

tetris-rl

Final project for Computational Sensorimotor Learning (6.484) at MIT. An agent for two-player Tetris trained using (unsupervised) reinforcement learning.

Gameplay samples: here, and here

Project Writeup