~/WRITING

Field notes.

Ideas, experiments, and the occasional rabbit hole.
A running log of what I’m building and learning.

01 ENTRIES / ALWAYS LEARNING

1 post tagged Deep LearningClear filter ↗

Multilayer perceptron diagram showing two input neurons, two hidden neurons, two output neurons, and the forward and backward flow of computation.

Oct 13, 2026 17 min read

Understanding Multilayer Perceptrons: Forward Propagation and Backpropagation

Learn how an MLP predicts and learns, with a complete numerical example covering forward propagation, output and hidden-layer gradients, weight updates, and practical limitations.

Machine LearningNeural NetworksBackpropagationDeep Learning