Weixi Tong

I am a Computer Science Ph.D. student at Purdue University, where I am advised by Prof. Tianyi Zhang.

I study how AI agents can carry out complex tasks and how people can understand, evaluate, and guide their work.

Portrait of Weixi Tong

Research Interest

My goal is to make AI agents dependable collaborators for complex, real-world work, with people able to understand and guide the process. I pursue this goal through three directions:

Agent capability
Enabling agents to plan and act effectively on complex tasks.
Agent evaluation
Assessing agent capabilities and diagnosing failures in code and interactive tasks.
Human–agent collaboration
Designing interactive systems that help people understand and steer agents.

Publications

Google Scholar

Preprints

Experience

Lilly Purdue Innovation Institute

Dec. 2025 – Present

Research Assistant · Purdue University

AI for science, agent self-improvement, and evaluation. Building agents with reusable skills and evaluation infrastructure for scientific computing.

Bosch Research USA

May – Aug. 2026

Research and Development Intern · Sunnyvale, CA

LLM and agentic AI research on tool use, policy adherence, and agent memory.

Education

Purdue University

2025 – Present

Ph.D. in Computer Science

Huazhong University of Science and Technology

2021 – 2025

B.Eng. in Computer Science and Engineering