path IV · Artificial Intelligence

Artificial Intelligence

Build learning from nothing: gradient descent, classical ML by hand, a neural net and an autograd engine you write yourself, up to a tiny transformer. No magic — just math you can see run.

0/9 subjects · each a complete document + a project
  1. 01
    What Learning Issoon

    Loss, gradient descent, and regression from scratch

  2. 02
    Classical Machine Learningsoon

    Logistic regression, trees, SVMs, and honest evaluation

  3. 03
    Unsupervised Learningsoon

    Clustering, PCA, and finding structure with no labels

  4. 04
    Neural Networkssoon

    The neuron, the forward pass, and backpropagation by hand

  5. 05
    Autograd & Trainingsoon

    Build your own micrograd and train it

  6. 06
    Deep Learningsoon

    Convolutions, sequences, and why nets actually train

  7. 07
    Transformers & LLMssoon

    Attention, a tiny language model, and how LLMs are trained

  8. 08
    Reinforcement Learningsoon

    Agents that learn to act: MDPs, Q-learning, policy gradients

  9. 09
    Generative Models & Alignmentsoon

    Diffusion, evaluation, failure modes, and where the field is going