Computational Intelligence, Control & Information Lab

We build agents that learn to perceive, act, and acquire skills in complex physical systems — uniting reinforcement learning, control and dynamical-systems theory, and information theory, with intrinsic motivation as a recurring thread.