HKUST
Hong Kong SAR, China
B.Eng. @ College of CS & Tech, Zhejiang University (ZJU)
CSE PhD @ Hong Kong University of Science and Technology (HKUST)
My research focuses on building autonomous agentic systems driven by foundation models (LLMs / MLLMs). Specifically, my work aims to advance the following areas:
Pushing capability upper bounds of agents and enabling continuous learning
Evaluating boundaries and reliability of agents
Grounding agentic systems in complex real-world scenarios
Throughout my journey in research, I've been fortunate to receive guidance from exceptional and supportive mentors. Currently, I am a PhD student at MMSense Lab, HKUST, advised by Prof. Yi R. (May) Fung. Previously, I was a research intern at SuDIS Lab, under the supervision of Prof. Huan Li and Dr. Lingxi Cui.
>> Please feel free to contact me if you are interested in discussing or collaborating with me. :D
News
- [2026.09] I have joined HKUST MMSense Lab as a PhD student, advised by Prof. Yi R. (May) Fung. ๐
- [2026.05] XSkill has been accepted by ICML 2026, see you in Seoul! ๐
- [2026.03] Two multimodal agent works, "XSkill" and "AgentVista", have released their papers and code! ๐
- [2025.05] Our paper "TableCopilot" was accepted by VLDB 2025! ๐
Education
Hong Kong University of Science and Technology
Ph.D. in Computer Science and Engineering
Zhejiang University
Undergraduate in Computer Science and Technology
The Affiliated High School of SCNU
High School Student
Research Experience
Research Intern @ Ren.AI Lab, HKUST
Advisor: Prof. Yi R. (May) Fung, Dr. Zhaochen Su
- Working on improving reasoning ability of multimodal agents, under the paradigm of "Thinking with Images".
Research Intern @ SuDIS Lab, ZJU
Advisor: Prof. Huan Li, Dr. Lingxi Cui
- Worked on discovery and management of structured data using LLMs and MLLMs.
Publications
XSkill: Continual Learning from Experience and Skills in Multimodal Agents
ICML26Guanyu Jiang *, Zhaochen Su *, Xiaoye Qu, Yi R. (May) Fung (*equal contribution)
AgentVista: Evaluating Multimodal Agents in Ultra-Challenging Realistic Visual Scenarios
PreprintZhaochen Su, Jincheng Gao, Hangyu Guo, Zhenhua Liu, Lueyang Zhang, Xinyu Geng, Shijue Huang, Peng Xia, Guanyu Jiang, Cheng Wang, Yue Zhang, Yi R. (May) Fung, Junxian He
TableCopilot: A Table Assistant Empowered by Natural Language Conditional Table Discovery
VLDB25Lingxi Cui, Guanyu Jiang, Huan Li, Ke Chen, Lidan Shou, Gang Chen
Industrial Experience
Structure-Guided Image Generation @ HunyuanDiT, Tencent
Advisor: Researcher Zhimin Li, Researcher Xinlu Lai
2024.07 - 2024.10
- Enhanced the architecture for structure-guided image generation by introducing a VLM with ControlNet training solution, receiving the Tencent Outstanding Student Award for contributions.
Services & Leadership
- Reviewer of AAAI2026 2025
- Deputy Secretary, Youth League Committee of Yunfeng, ZJU 2024 - 2025
- Member, "Qizhen Wenxue" Talent Researcher Program, ZJU 2024 - 2025
- Volunteer, 19th Asian Games in Hangzhou 2023
Beyond Research
>> Something about Guanyu.
Thoughts
Coming soonโฆ
