About Me

I am a third-year undergraduate student majoring in Computer Science at Xi’an Jiaotong University (XJTU). I recently completed an exchange program at UC Berkeley through the Berkeley Global Access (BGA) program in Spring 2025.

Currently, I am a research intern at Princeton AI2 Lab, proudly supervised by Postdoc Shilong Liu & Prof. Mengdi Wang, working on autonomous agents and GUI visual grounding. I am also collaborating with researchers at Peking University and the Beijing Innovation Center of Humanoid Robotics on world models for embodied agents.

Research Interests

I am broadly interested in building intelligent agents that can autonomously learn, reason, and act in complex environments. My current research focuses on:

• 🤖 Autonomous Agents: Self-evolving agents that learn transferable skills across diverse domains (terminal, web, structured data)
• 🌍 World Models: Scalable world models that enable embodied agents to plan and make decisions through imagination
• 🖥️ GUI Agents: Visual grounding methods for autonomous interaction with graphical user interfaces
• 🧠 Compositional Generalization: Learning reusable prototypes from limited data for generalization to unseen scenarios

I am passionate about developing AI systems that bridge the gap between passive observation and active interaction with the physical world.

News

• Dec 2025: Paper ZoomClick released — training-free GUI grounding via dynamic zooming
• Sep 2025: Paper WoW released — 14B world model trained on 2M robot trajectories
• Sep 2025: Started research internship at Princeton AI2 Lab
• Jan 2025: Began exchange semester at UC Berkeley

Selected Publications

ZoomClick: Unlocking and Evaluating the Potential of Zooming for GUI Grounding

Zhiyuan Jiang, et al. | arXiv 2025 | Paper | Code

WoW: Towards a World omniscient World model Through Embodied Interaction

Xiaowei Chi*, ..., Zhiyuan Jiang, et al. | arXiv 2025 | Paper | Project

Wow, wo, val! A Comprehensive Embodied World Model Evaluation Turing Test

Chun-Kai Fan*, ..., Zhiyuan Jiang, et al. | arXiv 2026 | Paper

View all publications →

Ongoing Projects

Autonomous Skill Learning for LLM Agents (Feb 2026 - Present)

• Building self-evolving agent framework with transferable skills across terminal, web, and structured data environments

Compositional Generalization of World Models (May 2025 - Present)

• Designing prototype-tree-based framework for extracting & reusing atomic features for zero-shot generalization

Automated AI-Researcher Agent (Jan 2026 - Present)

• Developing autonomous system for trend-aware research idea generation and manuscript writing

View all projects →

Education