Submitted by Polina Fedotova 322 Green-VLA: Staged Vision-Language-Action Model for Generalist Robots Sber Robotics Center 136 8