Publications

ReSo: A Reward-driven Self-organizing LLM-based Multi-Agent System for Reasoning Tasks
Heng Zhou, Hejia Geng, Xiangyuan Xue, Zhenfei Yin, Lei Bai
Annual Meeting of the Association for Computational Linguistics (ACL 2025, Under Review)
[Paper] [Code]

Position: Scaling LLM-based Multi-agent Systems Calls for Systematical Design
Guibin Zhang, Xiangyuan Xue, Yifan Hu, Sumeet Ramesh Motwani, Christian Schroeder de Witt, Lei Bai, Zhenfei Yin, Philip Torr
International Conference on Machine Learning (ICML 2025, Under Review)
[Paper]

ComfyBench: Benchmarking LLM-based Agents in ComfyUI for Autonomously Designing Collaborative AI Systems
Xiangyuan Xue, Zeyu Lu, Di Huang, Zidong Wang, Wanli Ouyang, Lei Bai
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2025)
[Paper] [Project] [Code]