Teaching

Diffusion Models: From Theory to Practice (6.S982)

Spring 2025 · CSAIL, MIT

Co-designed and co-taught this graduate MIT class with Prof. Costis Daskalakis. Class website.

Syllabus

  1. Introduction to generative models and their applications (GANs, VAEs, Flows, Diffusion Models, and Inverse Problems).
  2. Deep dive into Diffusion Models: the forward process, Itô integral, Itô formula, FP equation, reversibility, deterministic samplers, Tweedie’s formula, Denoising Score Matching.
  3. Diffusion models discretization error analysis.
  4. Learning diffusion models from corrupted data; likelihoods and latent diffusion.
  5. Flow Matching.
  6. Diffusion models and inverse problems.
  7. Schrödinger bridges.

Advanced Machine Learning: Topics in Unsupervised Learning

Spring 2023 · ECE, UT Austin

Teaching assistant for this graduate UT Austin class, where I also gave several lectures.

An advanced class focusing on topics in unsupervised learning, covering classical and modern techniques for modeling high-dimensional distributions: directed and undirected graphical models, learning graphical models using feature selection, submodularity, polytopes and combinatorial optimization, and inference in graphical models. Deep generative models covered include Variational Autoencoders, Generative Adversarial Networks, Autoregressive models, Normalizing Flows, Diffusions, and score-based models. We also explore problems in learning causal models and adversarial attacks.