Scholar
Sergey Levine
Google Scholar ID: 8R35rCwAAAAJ
UC Berkeley, Physical Intelligence
Citations & Impact
Citations
179,193
H-index
179
i10-index
558
Publications
20
Co-authors
184
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Developed end-to-end deep neural network policy training algorithms that integrate perception and control
- Proposed scalable inverse reinforcement learning algorithms
- Contributed key offline reinforcement learning algorithms including CQL, AWAC, and IQL
- Co-developed the D4RL benchmark suite
- Published influential work on real-world reinforcement learning at top venues such as NeurIPS
- Delivered numerous invited talks on robotic learning, safe RL, and planning with RL
