Iterative Compositional Data Generation for Robot Control

ArXi:2512.10891v4 Announce Type: replace-cross Collecting robotic manipulation data is expensive, making it impractical to acquire nstrations for the combinatorially large space of tasks that arise in multi-object, multi-robot, and multi-environment settings. While recent generative models can synthesize useful data for individual tasks, they do not exploit the compositional structure of robotic domains and struggle to generalize to unseen task combinations.