This is a type of recurrent neural network. Where each perseptron are connected to other perseptrons.
This Hopfield Network also introduced a new theory of Memory. [^1]
This is also connected to the problem of 3x3 connected node game. where I was treing to find the [[Modularity]] .
In this since the weight of the connection creats a gravity(force) on the state to get converge to a stable state which we call memory. That perticual thing answars may question.
[[Error correction code and modularity]] in this we have subspace that maps to the same value. where here we have a sub space which will converge to the same memory. [^1]
We can see Hopfield generates modulary but it has some flavor of [[Force]]. [^1]
Reference:
- [^1]Layerwise Lectures (Director). (2022, August 16). How are memories stored in neural networks? | The Hopfield Network. https://www.youtube.com/watch?v=piF6D6CQxUw