Education
Master at The University of Osaka under Prof. Chuan Xiao; bachelor's at HUST.
Master Student · The University of Osaka
Zichuan Xu.
Building systems at the edges of reinforcement learning, emergent language, and applied AI agents.
About
I am a master's student in the Graduate School of Information Science and Technology at The University of Osaka, advised by Prof. Chuan Xiao.
My research sits at the intersection of reinforcement learning and emergent language — how agents can develop, share, and adapt their own communication protocols under task pressure. On the engineering side I build and ship AI-agent systems for production: backends, tool-calling layers, and inference plumbing on constrained hardware.
Before Osaka I completed my bachelor's at Huazhong University of Science and Technology, and spent stretches working on graph learning for system-provenance security (at Virginia Tech, with Prof. Peng Gao), database tooling (Dameng), and full-stack web at Rakuten. I'm comfortable moving between research code and production code — and I think the second half of the 2020s will reward people who can.
— Updated May 2026
Research interests
What I'm thinking about.
Currently: how agents under task pressure invent and share their own communication, and how those signals stay stable under distribution shift.
Tech stack
What I build with.
Comfortable across the research → production gap — from PyTorch experiments to Laravel apps to Airflow pipelines and tool-calling backends.
Get in touch
Open to research collaborations and AI / agent-systems internships.
If you're working on emergent communication, multi-agent RL, or shipping agentic systems into production — I'd like to hear about it.
xu.zichuan@ist.osaka-u.ac.jp