Deep-dives on transformers, autograd, and LLM math — read the blog
Start to finish
Ordered reading paths. Each part assumes the one before it, so start at the top.
The whole path in order — linear algebra intuition, tokenizers, embeddings, attention, the full transformer forward and backward pass, optimizers, and two complete builds (autograd in NumPy, GPT-2 in PyTorch). Each part assumes the one before it.