Jun 22, 2025 Hi Robot: Open-Ended Instruction Following with Hierarchical Vision-Language-Action Models