Submitted by Polina Fedotova 237 Green-VLA: Staged Vision-Language-Action Model for Generalist Robots Sber Robotics Center 32 6