Skip to content
Clawdemy

References: Generating by denoising: diffusion

Source material

Section titled “Source material”
Source curriculum (structural mirror, cited as further study):
• MIT 6.S191, "Introduction to Deep Learning", Lecture 4: "Deep Generative Modeling"
  Instructors: Alexander Amini and Ava Amini (MIT)
  Course page: https://introtodeeplearning.com
  Code and labs: https://github.com/aamini/introtodeeplearning
  License: MIT (slides, code, and labs); videos are YouTube standard
  Required attribution: "© Alexander Amini and Ava Amini, MIT 6.S191:
    Introduction to Deep Learning, IntroToDeepLearning.com"
This lesson mirrors the diffusion portion of the generative-modeling lecture (the
VAE and GAN portions are in lesson 6). Clawdemy's lessons are original prose that
follows the pedagogical arc of this course. We do not reproduce or transcribe the
lectures; we cite them as the recommended companion. Course materials are used
under their MIT license with the attribution above; all rights to the original
videos remain with the creators.

Watch this next

Section titled “Watch this next”
  • MIT 6.S191, Lecture 4: Deep Generative Modeling by Alexander and Ava Amini. The lecture this lesson mirrors. Its generative-modeling material places diffusion alongside the VAE and GAN, so you can see the three approaches side by side with the instructors’ framing.

Going deeper

Section titled “Going deeper”

A short, durable list. Each link is a specific next step, not a generic pile.

  • “Denoising Diffusion Probabilistic Models” (Ho, Jain, and Abbeel, 2020). The paper that made diffusion practical and kicked off the modern wave of diffusion image generators. The primary source for this lesson; technical, but the forward-noising and reverse-denoising structure described here is exactly its backbone.

  • What are Diffusion Models? (Lilian Weng). A widely recommended written walk-through that bridges the intuition here to the actual math, the forward process, the reverse process, training, and conditioning, in careful steps. The best next read once “remove a little noise, many times” feels solid and you want the equations.

Adjacent topics

Section titled “Adjacent topics”

Where this connects inside the track.

  • Teaching machines to imagine (lesson 6). The previous lesson built the VAE and the GAN, the classic generative designs. Diffusion is the third route, and reading the two lessons together gives you the full map of how networks generate.

  • Learning by trial and reward (lesson 8). Every model so far, generative or not, learned from a fixed pile of data. The next lesson opens the final phase with a different kind of learning entirely: agents that learn by acting in an environment and improving from rewards.