Muning Wen
Papers
2
Total Citations
10
H-Index
2
About
Muning Wen is a rising researcher at the intersection of safe multiagent reinforcement learning (MARL) and large language model (LLM) agents. His work addresses two critical frontiers in AI: ensuring safety in multi-robot systems and enhancing the decision-making capabilities of LLM-powered agents. In his 2024 paper "Safe Multiagent Learning With Soft Constrained Policy Optimization in Real Robot Control," Wen tackles the urgent challenge of safety in MARL—a domain where few studies have ventured—by proposing a novel constrained optimization framework that enables multiple robots to learn cooperative behaviors while respecting hard safety limits, demonstrated in real-world robotic control. This work has already garnered 5 citations, signaling its timely importance. Complementing this, his paper "TRAD: Enhancing LLM Agents with Step-Wise Thought Retrieval and Aligned Decision" introduces a retrieval-augmented reasoning method that allows LLM agents to access relevant past reasoning steps, dramatically improving their performance on complex sequential tasks like web navigation. By bridging safety-critical MARL with more capable LLM agents, Wen is shaping the next generation of autonomous systems that are both powerful and trustworthy.
Research Focus
Key Achievements
Top Papers
- 1
- 2