Convolutional Neural Networks: CNN Math Explained
How do convolutional neural networks work? CNN math explained with kernels, padding, stride, pooling, receptive fields, and a full backpropagation example.
GPT Math Explained: The Full Forward Pass Beyond Attention
How does GPT math work end to end? One training step from token IDs through Q/K/V, attention, MLP, logits, cross-entropy loss, backprop, and AdamW.
Numerical Gradient Checking: Debug Your Autograd Engine
What is numerical gradient checking? The central difference formula, why you never train with finite differences, and a grad checker that catches real bugs.
Transformer from Scratch: Forward Pass and Backprop by Hand
One full transformer training step worked by hand: embeddings, positional encoding, attention, layer norm, cross-entropy loss, backprop, and Adam.
Backpropagation from Scratch: Build an Autograd Engine
How does backpropagation work? Build a working autograd engine from scratch in ~80 lines of Python: computation graphs, chain rule, reverse-mode autodiff.