As Scaling Law approaches its limits, reinforcement learning offers a new path to building more powerful large models. In the field of robotics, reinforcement learning has also brought about significant breakthroughs, exemplified by Spot, the robotic dog from Google's Boston Dynamics. With its exceptional flexibility and task execution capabilities, Spot has repeatedly hit the headlines and become the focus of public attention.