Quick references
Cheat Sheets
Essential machine-learning formulas, curves, and practical references. Each cheat sheet is a standalone Marimo notebook.
01→02→03→04→05→
Activation Functions
The main activation functions, shown one by one with their formula, curve, and use case.
Deep Learning Essential Formulas
A compact map of neural-network formulas from the forward pass through backpropagation, optimization, convolution, and attention.
Derivatives & Gradients
Reference sheet for basic, second, partial, and composite derivatives, with one short example per rule.
Log Loss, Bernoulli, and Sigmoid
Essential formulas connecting Bernoulli labels, likelihood, sigmoid probabilities, and binary log loss.
Loss Functions
Reference sheet for regression, classification, segmentation, and metric-learning losses, with gradients and short examples.