Generative Modelling Studio

Learn to make models that invent

Build intuition for VAEs, GANs, normalizing flows, and diffusion-style thinking. Expect animations, micro-quizzes, and the occasional AI dad joke to keep neurons firing.

Start with Introduction 🎛️ Interactive labs 📚 Research-inspired
🎨

From noise ➜ structure

Sampling, likelihoods, and playful experiments.

📚

Syllabus

Chapter 0 · Introduction

🔥

A friendly tour of explicit vs implicit models, a whirlwind of applications, and why safety matters. Includes quizzes, animations, and a tiny poem.

Jump in

Chapter 1 · Divergence Measures

📏

KL, JS, Wasserstein, TV, and $f$-divergences with interactive quizzes, true/false, and mini-labs.

Explore

Chapter 2 · Deep Learning

🧠

Perceptrons to Transformers—history, feedforward, backprop, batch norm, RNNs/LSTMs, masked attention, quizzes, and poems.

Dive in

Chapter 3 · Probabilistic Graphical Models

🧩

Factorization, conditional independence, hidden/observed setups, inference questions, quizzes, and memory poems.

Explore

Chapter 4 · Gaussian Mixture Models

🥣

Mixture basics, MLE + EM updates, variational inference, code demo, quizzes, and poems.

Explore

Chapter 5 · GANs

🎭

Generator vs discriminator, JS view, EM-style training loop, optimal D, stabilization tips, quizzes, and poems.

Explore

Chapter 6 · Wasserstein GANs

🌊

KR duality, 1-Lipschitz critics, WGAN-GP, full proofs, spot-the-mistake checks, and interactive quizzes.

Deep dive

Chapter 7 · Computer Vision Problems

👁️

Low & high-level vision, texture synthesis, denoising, quilting, NPR, and classical ideas that shaped modern generative models.

Explore

Chapter 8 · GAN Architectures for Vision

🏗️

Conditional GANs, Progressive training, Pix2Pix, StyleGAN, GigaGAN, GauGAN—specialized architectures for vision tasks.

Explore

Chapter 9 · Optimization Solvers

⚙️

Gradient descent, GDA, Nash equilibrium, game theory, and advanced solvers for min-max optimization.

Explore

Chapter 10 · Sinkhorn Generative Modeling

🚚

Optimal transport, entropy regularization, Sinkhorn algorithm, matrix scaling, and fast generative training.

Explore

Chapter 11 · Variational Autoencoders

🎲

Autoencoders, PCA connection, probabilistic latent spaces, reparameterization trick, ELBO, and generative modeling.

Explore

Chapter 12 · VAE Mathematical Theory

🎓

Divergence minimization, variational inference, complete proofs, KL for Gaussians, reparameterization trick, and posterior collapse.

Explore

Chapter 13 · Diffusion Models

🌊

Hierarchical VAE, VDM, DDPM algorithms, score matching, Langevin dynamics, and denoising score matching with complete derivations.

Explore

Chapter 14 · Optimization Methods

📈

Gradients, Jacobians, level curves, gradient descent/ascent algorithms, Cauchy's original 1847 paper, and convergence guarantees with complete proofs.

Explore

Explicit Density Models

🧮

Likelihoods, VAEs, and flows that keep math honest.

Coming soon

Implicit Density Models

🤝

GANs as creative rivals plus diffusion thinking and autoregressive tricks.

Coming soon

Evaluation & Safety

🛡️

Metrics, human evals, and building responsibly.

Coming soon

How this course feels

  • Short bursts of theory, followed by playful experiments.
  • Animations that show sampling, density shaping, and mode collapse.
  • Micro-quizzes with instant feedback (and pretend internet points).
  • Field notes from medicine, sports, entertainment, and education.
  • Safety interludes so we keep the creativity kind.

Mini mantra

"Sample boldly, evaluate kindly, repeat with a grin."

Yes, there will be the occasional rhyme. Generators like rhythm.

Ready to jam with probabilities?

Start with the Introduction chapter and unlock the first set of interactive labs.

Go to Introduction