A Hinton diagram in general represents vectors or matrices by representing each element with a square. The color of the square represents the sign of the element. In this case, green represents a positive number and red represents a negative number. The size of the square is based on the magnitude of the element. This means larger numbers will have larger squares representing them. The size is just relative to the size of every element and the scale between elements can be changed with the 'root' property of the function.
Example:
M =
1 -1 3
4 2 -1
-3 -2 1
Would have a Hinton diagram as seen in the PNG attached.
"plotwb" plots a Hinton diagram of weight and bias values. This means it takes the all the vectors and matrices used in network for the different layers and biases and plots them as Hinton diagrams. Each section of the diagram is one vector or matrix. For example the top left section represents the input vector and then section (2,2) represents the matrix that takes you from Layer 1 to Layer 2. Each section is labeled as from a source to a layer.
