Zhenjie Wan
CS undergrad @ Zhejiang University

Hi! This is Zhenjie Wan, a CS undergrad at Zhejiang University. I’m currently instructed by Prof. Wenguang Wang
. My research is centered around LLM/VLM as planners for embodied agents in 3D environments. I’m now interested in exploring how multi-modal LLM agents can collaborate with humans as equal partners in real-world tasks.
My dream is to develop a machine that can perceive, reason, and plan in real-world scenes like humans, and I have set this as the goal of my research. Toward this goal, I now pursue an open-ended embodied agent that can actively interact with humans and environments with large multi-modal models, particularly Large Language Models (LLMs) and Vision Language Models (VLMs), as its planner to do the reasoning. In my work, I study the sandbox game Minecraft as the environment for agents, as it is an open-world setting filled with long-horizon challenges, environmental disruptions, and uncertainties, closely resembling the human world in many aspects.
Specialties: Pytorch/CUDA, Collaboration, Basketball