Research & Projects

Published papers, open-source projects, and tools.

Publications

Qwen-CUA 2026

Qwen-CUA: Native Computer Use for (almost) Everything

Dunjie Lu, Shuai Bai, Tianyi Bai, Sicheng Fan, Chang Gao, Jian Guan, Feng Hu, Mianqiu Huang, Xingyang Huang, Yizhen Jiang, Yuheng Jing, Dehui Kong, Ning Li, Dayiheng Liu, Shixuan Liu, Zheng Liu, Que Shen, Bowen Wang, Junli Wang, Chencan Wu, Rui Xie

Technical Report, 2026

Computer-Use AgentGUI AgentReinforcement LearningQwen
arXiv:2608.02352
EvoCUA-1.5 2026

EvoCUA-1.5: Online Reinforcement Learning for Multi-turn Computer-Use Agents

Mianqiu Huang, Taofeng Xue, Chong Peng, Jinrui Ding, Sicheng Fan, Jiale Hong, Yufei Gao, Xiaocheng Zhang, Linsen Guo, Xin Yang, Dengchang Zhao, Yuchen Xie, Peng Pei, Xunliang Xie, Xipeng Qiu

Technical Report, 2026

Computer-Use AgentOnline RLGUI AgentLongCat
arXiv:2607.09773
CVPR 2026

WebChain: A Large-Scale Human-Annotated Dataset of Real-World Web Interaction Traces

Sicheng Fan, Rui Wan, Yifei Leng, Gaoning Liang, Li Ling, Yanyi Shang, Dehan Kong

IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026

GUI AgentDatasetWeb AgentMulti-modal
arXiv:2603.05295Code
ICLR 2026

WebFactory: Automated Compression of Foundational Language Intelligence into Grounded Web Agents

Sicheng Fan, Qingyun Shi, Shengze Xu, Shengbo Cai, Tieyong Zeng, Li Ling, Yanyi Shang, Dehan Kong

International Conference on Learning Representations, 2026

GUI AgentReinforcement LearningLLMSynthetic Data
arXiv:2603.05044Code

Open Source

WebChain

Active

WebChain is the largest open-source dataset of human-annotated trajectories on real-world websites, comprising 31,725 trajectories with 318,000 steps. First-author paper at CVPR 2026.

PythonDatasetGUI AgentCVPR 2026

WebFactory

Active

WebFactory is an automated RL training pipeline for GUI web agents, eliminating unsafe live web interactions and expensive human-annotated datasets. First-author paper at ICLR 2026.

PythonRLGUI AgentICLR 2026

WebClone

Active

Web Agent evaluation environment. Provides offline controllable website cloning for reproducible AI Agent testing, supporting batch data generation and standardized evaluation.

JavaScriptWeb AgentEvaluation

CafeMeet

Active

Smart meeting spot recommendation system. Using AI and map data analysis to intelligently recommend the best cafés for group meetings, considering ratings, distance, ambiance, and transportation.

PythonFastAPIAIMapOpenManus
Code
215