Zhouquan Liu

Papers

Total Citations

H-Index

About

Zhouquan Liu is a researcher at the forefront of interpretable artificial intelligence, with a focused expertise in bridging the gap between reinforcement learning (RL) and modular policy design. His primary research areas center on the development and optimization of Behavior Trees (BTs) for autonomous agents, particularly in robotics and computer games. Liu’s major contribution lies in pioneering methods to make RL algorithms more transparent and comprehensible. He has advanced the field by introducing evolution-inspired approaches to automatically learn and construct BTs, addressing the critical challenge of manually designing complex, error-prone AI policies. His work on "Interpretable Reinforcement Learning of Behavior Trees" directly tackles the significant AI challenge of interpretability, enabling developers to visually design and understand agent behaviors. With each of his most-cited papers accumulating 3 citations, Liu is establishing a foundational impact in a niche yet vital area. His research is particularly notable for its practical application, offering solutions that allow for reactive, modular, and task-switching capabilities in autonomous systems without requiring extensive expert domain knowledge.