Papers
3
Total Citations
57
H-Index
3
About
Sungbin Lim is a leading researcher at the intersection of reinforcement learning, robotics, and decision-making under uncertainty. His work is distinguished by pioneering theoretical advances in Monte Carlo tree search (MCTS) and entropy-regularized reinforcement learning, with direct applications to continuous control and soft robotics. Lim’s highly cited 2020 paper on MCTS in continuous spaces introduced Voronoi optimistic optimization, providing the first regret bounds for planning with discontinuous objectives—a breakthrough for domains like robotics and data-center management (29 citations). He also developed Generalized Tsallis Entropy Reinforcement Learning, a novel framework that unifies and generalizes maximum-entropy RL through an entropic index, enabling more flexible policy optimization for soft mobile robots (20 citations). In uncertainty-aware learning from demonstration, Lim proposed a sampling-free variance estimation method using mixture density networks, allowing robots to model complex, noisy human behaviors with a single forward pass (8 citations). His contributions bridge rigorous theory and practical deployment, making him a key figure in advancing sample-efficient, uncertainty-aware autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3