Your question is Deep Dive Into Diffusion and RL. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
Explain how you would approach diffusion policy and reinforcement learning for manipulation and mobile manipulation, including the pros and cons of different action spaces.