Comment by Affric

My guess would be that there’s a greater absolute magnitude of the vectors to get to the same point in the knowledge model.