Visual-language-action models, robotic foundational models, generalist robotic policy, deep reinforcement learning