Decompose and Recompose: Reasoning New Skills from Existing Abilities for Cross-Task Robotic Manipulation

ArXi:2605.01448v1 Announce Type: cross Cross-task generalization is a core challenge in open-world robotic manipulation, and the key lies in extracting transferable manipulation knowledge from seen tasks. Recent in-context learning approaches leverage seen task nstrations to generate actions for unseen tasks without parameter updates. However, existing methods provide only low-level continuous action sequences as context, failing to capture composable skill knowledge and causing models to degenerate into superficial trajectory imitation.