Jinghao Liu

I am an incoming MS student in Computer Science at UC San Diego, where I will work with Xiaolong Wang on robot learning. I recently received my BS degree in Computer Science from the Paul G. Allen School at the University of Washington.

My research interests are in robot learning and foundation models — in particular how a single model can generalize across many robots, many sensors, and many compute budgets.

I am always happy to talk about research. Email is the best way to reach me.

news

Aug 12, 2026	Flex-π, my first paper, is out on arXiv — project page.

selected publications

arXiv
Flex-π: A Multi-Stream World-Action Model with Compute Flexibility

Ge Yan^*, Jinghao Liu^*, Yuzhi Fan^*, Lei Cai, Minwen Liao, Jesse Zhang^†, and Dieter Fox^†

arXiv preprint arXiv:2608.10860, 2026

Abs arXiv Bib Website

One checkpoint. Any observed inputs, any generated futures, and a speed–accuracy operating point you pick at deployment.
@article{yan2026flexpi, title = {Flex-π: A Multi-Stream World-Action Model with Compute Flexibility}, author = {Yan, Ge and Liu, Jinghao and Fan, Yuzhi and Cai, Lei and Liao, Minwen and Zhang, Jesse and Fox, Dieter}, journal = {arXiv preprint arXiv:2608.10860}, year = {2026}, }