Submitted by Polina Fedotova 324 Green-VLA: Staged Vision-Language-Action Model for Generalist Robots Sber Robotics Center 120 8