Zhaopeng Meng

Papers

Total Citations

181

H-Index

About

Zhaopeng Meng is a researcher specializing in deep reinforcement learning (DRL) and multi-agent reinforcement learning (MARL), with a particular focus on improving the efficiency and applicability of these methods across complex real-world domains. His work spans critical challenges in AI, including autonomous vehicles, robotics, and game intelligence. Meng's most significant contribution is his comprehensive survey on exploration strategies in deep reinforcement learning, spanning both single-agent and multi-agent settings. This work, which has garnered an impressive 158 citations, addresses one of the field's most persistent bottlenecks — sample inefficiency — by systematically cataloging and analyzing exploration techniques that help agents learn more effectively with limited interaction data. An earlier version of this survey laid the groundwork for the widely cited iteration, reflecting the sustained relevance of his research agenda. Beyond survey work, Meng has tackled the underexplored challenge of hybrid action spaces through his HyAR framework, which bridges discrete and continuous action representations — a practical problem in robot control and game AI that most prior RL approaches had left unaddressed. Together, his publications demonstrate a consistent commitment to making reinforcement learning more scalable, efficient, and broadly applicable, establishing him as a thoughtful contributor to modern AI research.