Yunfu Deng

I am a robotics PhD candidate at the University of Wisconsin–Madison, advised by Prof. Josiah P. Hanna. My research focuses on robot learning under limited real-world interaction, with an emphasis on building models that help robots predict, evaluate, and improve their behavior beyond the data they were trained on.

My current work spans world models, sim-to-real transfer, and policy evaluation for vision-based robot policies. More broadly, I am interested in reliable robot decision-making: how robots can learn useful representations of the physical world, reason about long-horizon consequences, and identify when their learned models or policies may fail.