Kangye Ji

Kangye Ji is a Ph.D. student in computer science at Tsinghua University, affiliated with the Efficient Deep Learning and Embodiment Group under the supervision of Professor Zhi Wang. His research focuses on efficient Embodied AI, particularly the acceleration of vision-language-action models and diffusion policies for real-time robotic control. He is best known as the lead author of Block-wise Adaptive Caching (BAC) and Sparse ActionGen, two training-free methods for substantially speeding up Diffusion Policy inference in robotic manipulation tasks.