Sergey Levine

Sergey Levine is an Associate Professor at the University of California, Berkeley, affiliated with the Department of Electrical Engineering and Computer Sciences in the College of Engineering. His research focuses on autonomous decision-making systems, particularly in deep reinforcement learning, robotics, and computer vision.

• PhD, Computer Science, Stanford University (2014)
• BS/MS, Computer Science, Stanford University (2009)

Levine develops algorithms for end-to-end training of neural network policies, inverse reinforcement learning, and safe real-world RL. His work enables autonomous agents to acquire complex behaviors through learning frameworks.

Recent publications emphasize diffusion models, offline RL, and vision-language-action systems for robotics. Awards include NSF PECASE (2024), Sloan Fellowship (2019), and MIT TR35 (2016).

• NSF Presidential Early Career Award (2024)
• Okawa Research Grant (2021)
• Sloan Research Fellow (2019)
• NSF CAREER Award (2017)
• MIT Tech Review TR35 (2016)
• ONR Young Investigator (2016)

He leads the Robotic AI & Learning Lab (RAIL) at Berkeley Artificial Intelligence Research (BAIR), contributing to open-source frameworks like Octo and OpenVLA. His courses (CS 285, CS 182) cover deep RL and learning theory fundamentals.