How does this stack up to diffusion policy learning? https://diffusion-policy.cs.columbia.edu/ https://diffusion-policy.cs.columbia.edu/ Looking at the figure

How does this stack up to diffusion policy learning? https://diffusion-policy.cs.columbia.edu/ https://diffusion-policy.cs.columbia.edu/ Looking at the figures and videos it seems... worse? Sort of surprised they didn't compare it but I guess they're trying to limit the discussion to purely reinforcement learning methods. EDIT: Ah, I see this is an older result and was published concurrently to the diffusion policy paper, so it's likely the authors didn't know about it in time to add the extra comparisons.