It really does get it, because MTP is usually run at "3 token" depth. It's pretty shocking to watch

I think you’re missing that MTP can predict more than 1 token in advance.

Off the top of my head, I'm guessing we're missing sparse attention. But I'll run your challenge through and see where the gaps are. I promise I'm telling the truth :)