Sergey Levine works as a computer scientist and professor at the University of California, Berkeley. His research focuses on robotics and machine learning, and he helped start the company called Physical Intelligence.

Education

Levine earned a bachelor's degree and a master's degree in computer science from Stanford University. During college, he worked as an intern at NVIDIA. Levine finished his doctorate in computer science at Stanford, where his research studied how robots learn, ways to make robots move efficiently, and methods to use data to create instructions for robots in advanced systems. After earning his doctorate, Levine worked as a researcher at the Robot Learning Lab at UC Berkeley with Pieter Abbeel.

Academic career

In 2015, Levine became a part-time research scientist at Google to work on machine intelligence.

In 2016, Levine joined the faculty of the University of California, Berkeley, where he is a professor in the Department of Electrical Engineering and Computer Sciences. At Berkeley, he leads a research group that studies the connection between robotics, machine learning, and control.

His research focuses on reinforcement learning, imitation learning, and systems that help robots learn on a large scale. Levine’s work includes both model-based and model-free methods, with an emphasis on helping robots and other autonomous agents learn skills from large amounts of data. His group has developed techniques that use deep neural networks to directly connect complex sensory inputs, such as images, to actions that control movement.

Research contributions

Levine worked on advanced learning methods to help robots perform tasks. He created guided policy search, a technique that teaches computers to perform complex robot tasks. He also helped develop methods that allow robots to learn directly from visual information and motor actions, improve learning efficiency using models, and train using large robot data sets without real-time interaction. His research helps robots learn complicated behaviors from complex sensory information and apply these skills to different tasks and environments.

Awards

Sergey Levine has been awarded the Presidential Early Career Award for Scientists and Engineers (PECASE) in 2025, and the MIT Technology Review’s Innovators Under 35 in 2016.