one thing that's remarkable (and sadly far too rare in academia) is that the top papers come with working code. there's no dumb kneecapping of your competitors as we see in so many other fields.