Learn to use AI

How a network learns

How a network learns is a unit of 8 lessons in Rabbithole's Learn to use AI course, at level 2.

Lessons

1. Activation functions: the bend that lets networks learnLevel 2
2. Gradient descent: learning in small stepsLevel 2
3. Backpropagation: how a network finds what to changeLevel 2
4. Softmax: from raw scores to chancesLevel 2
5. Residual connections: the shortcut around each stepLevel 2
6. Hyperparameters and the learning rateLevel 2
7. Regularisation: keeping a model from memorisingLevel 2
8. Reading Learning CurvesLevel 2

Sources

These lessons draw on 15 sources. The most used:

• Stanford CS224N: Natural Language Processing with Deep Learning, Spring 2024
• Machine Learning Crash Course
• MIT 6.S191: Introduction to Deep Learning
• MIT 6.S191: Introduction to Deep Learning (2026)
• Hugging Face LLM Course
• Neural Networks: Zero to Hero
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Deep Learning

All units in Learn to use AI