About Me

I'm Sicheng Fan, a Master's student at Fudan University, working on GUI Agents/Computer-Use Agents and Reinforcement Learning. I also serve as the Technical Lead at WebAgentLab community and as an AI Research Intern at LongCat Group, Meituan.

My research focuses on building Computer-Use Agents, utilizing Vision-Language Models (VLMs) for cross-platform task automation, to achieve the next generation of Computer-Use Agents. Recent work includes WebChain (CVPR 2026, the largest web interaction trajectory dataset) and WebFactory (ICLR 2026, an automated RL training framework for GUI agents).

I actively contribute to open source and aim to advance the GUI agent field through shared research and tools. Feel free to reach out for collaboration!

Education

Master's Student

Fudan University

Sep 2024 — Present

Research: GUI Agent/Computer-Use Agent, Reinforcement Learning

Bachelor's Student

Fudan University

Sep 2020 — Jun 2024

Outstanding Graduate of Shanghai, Outstanding Student of Fudan University

Experience

AI Research Intern

LongCat Group, Meituan

Apr 2026 — Present

Participating in AI Research at LongCat Group, focusing on Computer-Use Agent research.

GUI Agent Researcher

iMeanAI

Jun 2024 — Present

Conducting GUI agent research, leading WebChain and WebFactory projects.

Technical Lead

WebAgentLab

2025 — Present

Leading technical direction, building the open-source GUI agent ecosystem.

Research Areas

GUI Agent

VLMComputer UseWeb AutomationScreen Understanding

Reinforcement Learning

PPORLHFReward ModelingPolicy Optimization

LLM Training & Fine-tuning

SFTDPOAlignmentInstruction Tuning

World Model

State PredictionPlanningSimulation

Computer-Use Agent

Computer UseWeb AutomationScreen Understanding

Contact

Technical Skills

PythonC/C++TypeScriptPyTorchTensorFlowReactNext.jsFastAPIGitDocker

Stats

4
Papers
2
Top Venues
6
Blog Posts
5+
Yrs Coding