• Stanford Percy Liang: Strong understanding of basics: We would like you to deeply understand the things that you have learned. We welcome candidates of all levels, so it is definitely not necessary to know the latest work in AI or have taken CS229. But if you have taken CS229, you should know logistic regression very well. Passion and drive: Unlike classes, research is unstructured, open-ended, and relies heavily on intrinsic motivation. Successful researchers are proactive about getting unstuck and just really want to make progress towards a goal without external pressure. Curiosity: Research is about asking good questions, challenging assumptions, and asking why you’re doing something, not just going through the motions. Reliability: Like anything else in life, research projects have deadlines that need to be met, and clear communication is important (especially if you cannot meet a deadline).

• CMU Deepak Pathak: Our group studies Artificial Intelligence at the intersection of Computer Vision, Machine Learning & Robotics. Our ultimate goal is to build agents with a human-like ability to generalize in real and diverse environments. We believe understanding how to continually develop knowledge and acquire new skills from just raw sensory data will play a vital role in achieving this goal. Our group draws inspiration from psychology to build practical systems at the interface of vision, learning and robotics that can learn using data as its own supervision. If you would like to join our group, please fill this form and then send me a short email note without any documents.

• Georgia Tech Xu Danfei: Examples for robotics include but not limited to: ROS, OpenRave, MoveIt, OMPL, Gazebo, Mujoco, pyBullet, IssacGym, AI2Thor, hand-eye calibration, multi-camera systems, pose estimation, SLAM, Task and Motion Planning, legged robot, arm manipulators, mobile manipulators, UAVs. Examples for ML and control include but not limited to: pytorch, tensorflow, RL, optimal control, imitation learning, generative modeling, distributed training, representation learning, 3D perception.