Danfei Xu is a Research Scientist with the Autonomous Vehicle Research Group. Prior to joining NVIDIA, Danfei obtained his Ph.D. in Computer Science from Stanford University in 2021, advised by Silvio Savarese and Fei-Fei Li. He obtained his B.S. from Columbia University in 2015. Danfei’s research is in robot learning. His research goal is to build autonomous systems that have human-like abilities to generalize to new tasks and environments. His Ph.D. work sought to endow robots with compositional generalization capabilities and solve long-horizon tasks in complex environments. Some examples are generalizable visual imitation learning using neural program induction and neural graph inference, compositional planning with neural-symbolic planners, learning abstract planning spaces, and discovering compositional representations from demonstrations.