The CAB Lab is broadly interested in developing neural network models with physical priors toward the goal of improving the interpretability and generalizability of neural network solutions. We consider physical priors broadly (e.g., matrix groups in machine perception and robotics problems, and symplectic geometry in dynamical systems) and embed them into neural network systems through architectural design. PI Allen-Blanchette co-appointed in the Department of Mechanical and Aerospace Engineering and the Center for Statistics and Machine learning, and is affiliated with the Department of Computer Science and Robotics at Princeton.