Heng Fang

Heng Fang is a Ph.D. student in the Department of Robotics, Perception and Learning at KTH Royal Institute of Technology, where he is advised by Prof. Hossein Azizpour, Prof. Andrea Nascetti, and Prof. Josephine Sullivan. He is the lead author of the paper "World In-Context Learning for Agentic Robot Manipulation," which introduces the WICL paradigm and the RoboScript infrastructure for agentic robotic manipulation. His broader research interests span deep learning, temporal modeling, uncertainty estimation, and their applications in areas including satellite imagery and biomedical scenarios.