-- Nov 10 In-Class Exercise Thread
We want to produce an output that is a number modulo 8.
These numbers are {0,1,2,3,4,5,6,7} which can be represented as 3-bit binary numbers.
- We will use 3 neurons where each neuron will represent the bits in the output.
- Every neuron will get 4 inputs (1 input bit from the sequence + 3 neuron outputs from the previous time step)
The following diagram will summarize the network.
Neurons are numbered 1 through 3 from top to bottom in the above diagram.
Neuron 1 will compute: `(y_1)^t = (y_1)^(t-1) OR ((y_2)^(t-1) AND (x^t AND (y_3)^(t-1))`
Neuron 2 will compute: `(y_2)^t = (y_2)^(t-1) OR (x^t AND (y_3)^(t-1))`
Neuron 3 will compute: `(y_3)^t = (x^t OR (y_3)^(t-1))`
(
Edited: 2021-11-15)
We want to produce an output that is a number modulo 8.
These numbers are {0,1,2,3,4,5,6,7} which can be represented as 3-bit binary numbers.
* We will use 3 neurons where each neuron will represent the bits in the output.
* Every neuron will get 4 inputs (1 input bit from the sequence + 3 neuron outputs from the previous time step)
The following diagram will summarize the network.
((resource:Screen Shot 2021-11-15 at 10.13.27.png|Resource Description for Screen Shot 2021-11-15 at 10.13.27.png))
Neurons are numbered 1 through 3 from top to bottom in the above diagram.
Neuron 1 will compute: @BT@(y_1)^t = (y_1)^(t-1) OR ((y_2)^(t-1) AND (x^t AND (y_3)^(t-1))@BT@
Neuron 2 will compute: @BT@(y_2)^t = (y_2)^(t-1) OR (x^t AND (y_3)^(t-1))@BT@
Neuron 3 will compute: @BT@(y_3)^t = (x^t OR (y_3)^(t-1))@BT@