About me
I am a fourth-year Ph.D. student in the Computer Science Department at Carnegie Mellon University and a member of the Foundations of Cooperative AI Lab, advised by Vincent Conitzer.
My research focuses on multi-agent LLM systems: how to make LLMs cooperate and coordinate effectively within a system, and how to keep such systems efficient, truthful, and economically sustainable. I combine LLM research with game theory and mechanism design to study how incentives and strategic behavior shape what LLM agents do. Recent directions include:
- Cooperation and coordination among LLM agents: how memory and post-training affect cooperation among LLM agents (Memory Curse), and how to keep multi-agent collaboration robust against adversarial majorities (Consensus Trap).
- Incentive-aware agentic systems: truthful model routing and agentic workflow design (Truthful Routing, EMNLP 2026).
- LLMs meet mechanism design: generative advertising in LLMs (COLM 2026) and automated mechanism design (EC 2025 Workshop, NeurIPS 2026).
Education. I received my undergraduate degree from the Institute for Interdisciplinary Information Sciences (Yao Class) at Tsinghua University, advised by Zhixuan Fang. In 2022, I visited the Harvard EconCS Group, hosted by Yiling Chen.
Industry Research Internship Experiences.
- 2026, ByteDance Seed-LLM (San Jose): LLM post-training and real-time agents.
- 2023, Microsoft Research, hosted by Siwei Wang.
- 2021, Shanghai Qizhi Institute, hosted by Zhixuan Fang.
I am always happy to discuss research and open to collaborations — feel free to reach out!
Contact: jiayuan4@cs.cmu.edu