Mike Hang Wang

소개

Mike’s research focuses on developing AI agents that learn and continually evolve through direct interaction with the physical world (e.g. human feedback). This vision is pursued through three interconnected research areas:

• Warm-start Reinforcement Learning and Self-supervised Learning,
• Multi-agent Reinforcement Learning (distributed optimization), and
• Foundation models (with particular emphasis on World Models).

His work spans both theoretical innovations and practical algorithmic implementations, with applications in LLM agents, robots, Internet of Things (IoT), and edge computing.