1 paper
Xunyao Lei, Renjun Wu, Tianlin Huo +1
At every re-observation point in a hierarchical Vision-Language-Action (VLA) system, two interface decisions must be made: when to terminate the current subtask and how far to exec…