Spatial Intelligence & Agentic AI
From Persistent Worlds
to Self-Evolving Agents.
I build agents that understand 3D worlds, act through world models, and improve themselves through experience.
See the research in action.
Three connected frontiers: perceive and preserve the world, use it for action, then improve the intelligence itself.
Persistent Worlds
Using RGB input only, HorizonStream delivers stable streaming 3D reconstruction beyond 10K frames, with constant memory and linear time.
Actionable World Models
Moving world models from plausible generation to controllable prediction and measurable decision utility.
AI that Improves AI
A unified view of AI systems that improve data, training, evaluation, workflows, and eventually themselves.
Selected highlights.
Research for open-ended worlds.
I am a tenure-track Assistant Professor in the AI Thrust at HKUST(GZ). My group works across spatial intelligence, multimodal learning, and agentic AI, from persistent 3D environments to self-improving agents.
Build intelligent worlds together.
Open to joint research and industry collaboration in 3D spatial computing, digital twins, embodied and game agents, generative AI, and self-evolving systems.