Georgia Tech · PhD Student

Jiyi Wang.

I work on robot learning and embodied AI: adapting vision-language-action models with human feedback. I also work on discovering reusable motor skills from animal behavior data.

At Georgia Tech, I work with Dr. Anqi Wu and Dr. Ye Zhao on dexterous manipulation. My research bridges real-world robotic control and reinforcement learning. Previously, I was working on brain-inspired AI and building more reliable learning systems.

Download CV Email
Google Scholar ↗ GitHub ↗

Robot learning

I fine-tune vision-language-action policies with human-in-the-loop reinforcement learning for precise robotic manipulation. Iterative hardware RL improved millimeter-level task success from 10% to 85%.

Continuous pipette alignment

17 sec

A real-robot demonstration of repeated pipette alignment over a tip rack.

Open video ↗

Responding to perturbations

8 sec

Pipette alignment while a person moves the tip rack during execution.

Open video ↗

Selected publications

All publications

Seeking 2027 research internships in robot learning and embodied AI. jwang3617@gatech.edu