Introduction to Week 7
This week we’re going to begin our pivot from LLMs to other forms of generative AI by learning about diffusion models. This week will be like weeks 1 and 2, where the explainers we’ll watch contain some math and some computer code, but we’re not trying to learn the math or the code – our goal is to develop an understanding of the key concepts underlying diffusion models.
Weekly Activity: Part 1
For the first part of this week’s activity, watch the following videos. This will take just under an hour, so leave yourself plenty of time. These videos will introduce you to key concepts associated with diffusion models. (The second video says it’s explaining the concept at four different levels of difficulty, but it’s actually explaining it in steps. Don’t let that confuse you.) As always, repetition is your friend here, so watch all the videos even though they cover the same topics. And as always, you can ignore the specifics of the math and the code shown in the videos. You’re just trying to develop a solid conceptual understanding of a few key ideas. Listen for information that will help you answer the following questions:
- What does it mean to diffuse noise into an image (in other words, add noise to an image)?
- What does it mean to remove noise from an image?
- What role do neural networks play in diffusion models?
- How can text be used to steer a diffusion model? In other words, how do text-to-image models work?
Weekly Activity: Part 2
Start by reviewing any of this week’s key concepts you feel like you’re not understanding yet with your LLM. You might try asking your LLM to “explain it to me using _________ as an example” (anything you’re interested in). Ask additional questions to make sure you’re understanding.
Then copy and paste the prompt below into your LLM and answer the questions to complete the review. When you are finished, copy and paste the entire review conversation into Assignment 7: Diffusions Models. This assignment is due by 11:59pm Mountain on Thursday, February 22.
I've just finished studying about diffusion models. I'm trying to develop a solid conceptual understanding of the key ideas. Give me a quiz where you ask me the following questions:
- What does it mean to diffuse noise into an image (in other words, add noise to an image)?
- What does it mean to remove noise from an image?
- What role do neural networks play in diffusion models?
- How can text be used to steer a diffusion model? In other words, how do text-to-image models work?
Ask me one question at a time and wait for my answer. After each answer, give me feedback on my answer and explain anything it seems like I don't understand. Then ask if I'd like additional information on that question. When I indicate I'm finished, ask me the next question.