Last active
February 16, 2026 21:04
-
-
Save kripken/9582513853a6b6be3a777331d52b5ce9 to your computer and use it in GitHub Desktop.
Greeting Neural Network
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| def matrix_math(matrix, offset, vector): | |
| ''' | |
| Computes | |
| matrix*vector + offset | |
| Matrix is mxn, vector ix 1xn, offset is 1xn | |
| ''' | |
| m = len(matrix) | |
| n = len(vector) | |
| ret = [0] * len(offset) | |
| for i in range(m): | |
| for j in range(n): | |
| ret[i] += matrix[i][j] * vector[j] | |
| ret[i] += offset[i] | |
| return ret | |
| class NeuralNetwork: | |
| ''' | |
| Given a name in the input layer, emit "Hello, {NAME}" in the output. | |
| ''' | |
| max_name_len = 8 | |
| def linear_layer(self, input): | |
| ''' | |
| Copy the input into the right place, with the rest of the string in | |
| the offset. | |
| ''' | |
| weights = [ | |
| # First 7 spots are will be filled in by 'Hello, ' | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 0], | |
| # Next, we copy in the input. | |
| [1, 0, 0, 0, 0, 0, 0, 0], | |
| [0, 1, 0, 0, 0, 0, 0, 0], | |
| [0, 0, 1, 0, 0, 0, 0, 0], | |
| [0, 0, 0, 1, 0, 0, 0, 0], | |
| [0, 0, 0, 0, 1, 0, 0, 0], | |
| [0, 0, 0, 0, 0, 1, 0, 0], | |
| [0, 0, 0, 0, 0, 0, 1, 0], | |
| [0, 0, 0, 0, 0, 0, 0, 1], | |
| ] | |
| # 'Hello, ' + zeroes | |
| offset = [72, 101, 108, 108, 111, 44, 32, 0, 0, 0, 0, 0, 0, 0, 0] | |
| return matrix_math(weights, offset, input) | |
| # Run the network on some tests | |
| nn = NeuralNetwork() | |
| for name in ['Anna', 'Joe', 'Crevyzik']: | |
| # Convert name to numbers. | |
| input = [ord(c) for c in name.ljust(nn.max_name_len, ' ')] | |
| # Run the network. | |
| output = nn.linear_layer(input) | |
| # Convert back to letters , and print. | |
| output = ''.join([chr(i) for i in output]) | |
| print(f'input: {name.ljust(nn.max_name_len)} ==> ', end='') | |
| print(f'neural network output: {output}') |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment