Submitted by Polina Fedotova 319 Green-VLA: Staged Vision-Language-Action Model for Generalist Robots Sber Robotics Center 111 8