λFPAI

Learn / Deep Learning

Deep Learning

Neural networks as compositions of linear maps plus nonlinearities — and the training dynamics that make them work.

Overview

Introduction to deep learning: neural networks and training dynamics.

Neural networks

Perceptrons, activations, and representations.

Training

Backprop, optimizers, and stability.

CNNs

Convolutions, locality, and vision models.

Transformers

Attention, scaling laws, and sequence modeling.

Generative models

Diffusion, autoregressive models, and sampling.