Skipping levels in a hierarchy

I was thinking about the semantics thing and as I often do - I looked at how the brain does it. The map to map connections includes a significant number of connection that skip a level. See figure 2 in this paper and note the map “skipping” in semantic learning.

I have been chatting with @Paul_Lamb and @gmirey about this and we have come up with a few possibilities.
I want to throw this out to the community and see what you fellow theorists think about this.

One possibility is: I have been leaning towards the “top boss” idea that this upper level is a boss that learns quickly with the long loops first. As the “middle” layer learns the same input (with counter-flowing feedback from the boss) it adds its influence to the lower resolution input and the two signals together form a new input to the boss that may change it to a more nuanced interpretation of the input.

Another idea that was offered is: So perhaps the middle layer would provide some workspace to ease moving a frequently encountered concept down the hierarchy. The more informed higher level guy can provide smarter feedback based on having more information, then the middle guy would start to get the hang of it and replace the higher guy when he becomes more certain.


Another dynamic of connecting each level to the next two higher levels, is that it provides six links that can be tweaked with different properties, such as degree of influence and learning/forgetting rates. It may be beneficial for one of the higher levels to be optimized for rapid one-shot learning, at the cost of more rapid forgetting and object representations bleeding into each other, while the other level learns more slowly, forming more accurate and permanent representations.

1 Like

Including Jeff’s observations on this level-skipping phenomenon as well (in this case, the focus is on scale, which probably more applicable when talking about the lowest hierarchical levels closest to the sensors)