Bradly C Stadie
Assistant Professor of Statistics and Data Science
PhD, UC Berkeley
- bradly.stadie@northwestern.edu
- Website
- 2006 Sheridan Road, Room 101B
Research Interests
My research area is reinforcement learning (RL). In particular, I'm interested in using feedback to teach machines new behaviors. This is most commonly applied to robotics, biological sciences, and large language models.
Using techniques from RL, we were recently able to use ChatGPT to control a robot: https://wonderful-team- robotics.github.io/
I'm also interested in the intersection of RL and diffusion models. For example, here is a system we developed where a diffusion model and human jointly share control of a robot: https://diffusion-for-shared- autonomy.github.io/
My Google Scholar page can be found here.