Foundations

    Build the mathematical and practical foundations to understand, implement, and reproduce AI research.

    A path into research

    1. 01

      Learn the foundations

      Linear algebra, probability, and optimization.

    2. 02

      Implement a model

      Build a small model and trace how it learns.

    3. 03

      Reproduce a paper

      Match its setup before changing the method.

    4. 04

      Design an experiment

      Use a baseline and ablations to test your idea.

    Math Foundations4 resources

    3Blue1Brown Linear Algebra

    Visual intro to vectors and matrices

    3Blue1Brown Calculus

    Essence of calculus series

    Mathematics for ML

    Imperial College London textbook, free online

    Matrix Calculus for DL

    Derivatives and gradients explained for deep learning

    Courses & Fundamentals5 resources

    Andrej Karpathy: Zero to Hero

    Build neural networks and GPT from scratch, in code

    Stanford CS224n

    NLP with deep learning — the standard LLM foundations course

    Stanford CS231n

    CNNs for visual recognition

    3Blue1Brown Neural Networks

    Visual explanations of neural networks and transformers

    Deep Learning Book

    Goodfellow, Bengio, Courville textbook

    Architectures4 resources

    Attention Is All You Need

    The original Transformer paper (2017)

    The Illustrated Transformer

    Visual walkthrough of the Transformer architecture

    ResNet Paper

    Deep residual learning (2015)

    BERT Explained

    Understanding the BERT architecture visually

    Training & Optimization4 resources

    PyTorch Tutorials

    Official training guides

    Hugging Face Training

    Fine-tuning transformers in practice

    LoRA Paper

    Low-rank adaptation — the fine-tuning method everyone uses

    Adam Optimizer Paper

    Adaptive moment estimation (2014)