Reading order
Softmax
The softmax function turns scores (logits) into probabilities that sum to 1 by dividing each exp(logit) by the sum. Scores (2.0, 1.0, 0.1) become (0.659, 0.242, 0.099). Covers why exp, temperature, the sigmoid, nan, and the derivative.
Backpropagation
Cross-Entropy
Attention
Sigmoid
Top-k and Top-p Sampling
Activation Functions
Loss Function
Linear Regression
Vanishing Gradient
Automatic Differentiation
Softmax Regression
Modes of Machine Learning
Perceptrons
XOR and Multilayer Perceptrons
Expert System Rules
Big Data and Bias
Word Sets for Text
Bag of Words
School Model of TF-IDF
Text Similarity
Lexicon-Based Sentiment Analysis
Pixel Matrices
Image Brightness Transforms
Image Transposition
Image Classification with Hamming Distance
Prediction from Relative Frequencies
Prediction with Trend Lines
Univariate Squared Loss
Slope of a Loss Function
Gradient Descent in One Variable
AI Decision-Making Trade-offs
An AI Mathematics Inquiry