Sergey Levine

Sergey Levine is an Associate Professor in the Department of Electrical Engineering and Computer Sciences at UC Berkeley, where he leads pioneering research at the intersection of machine learning, decision making, and control. His work centers on developing algorithms that enable autonomous systems to learn complex behaviors directly from data, with a particular emphasis on deep reinforcement learning, inverse reinforcement learning, and scalable methods for end-to-end policy training that tightly integrate perception and action. Levine’s group has produced foundational contributions to robotic manipulation, vision-language-action models, autonomous decision processes, and model-based and offline RL frameworks. Applications of his research span autonomous robots and vehicles, simulation-to-real transfer, human-robot interaction, and large-scale learning systems. Before joining Berkeley in 2016, Levine received his BS, MS, and PhD in Computer Science from Stanford University. His lab collaborates broadly across robotics, computer vision, and AI theory, and continues to push the frontier of physically grounded intelligence and generalizable machine learning policies.