This is awesome. If you made a book or video-course that takes this level of high level explanation and translate it to the technical and then mathematical level, I would buy it in an heartbeat.
This is what I think is missing in most AI (broad sense) learning resources. They focus too much on the math that I miss the intuitive process behind it.
It’s purposefully high level and non-technical for a general audience - my theory was that most people who aren’t into tech/AI don’t care too much about training, or how the system got to be the way that it is.
But they do have some interest in how it actually operates once you’ve typed in a prompt.
Happy to answer any questions or take on board feedback
I think some of the visualizations would be much better if you used a pixel-space model instead of a latent diffusion model.
Right now we are only seeing the denoising process after it's been morphed by the latent decoder, which looks a lot less intuitive than actual pixel diffusion.
If you can't find a suitable pixel-space model, then you can just trivially generate a forward process and play it backwards.
The interpolations between butterfly and snail were pretty horrifying. But something like Z-Image you could basically concatenate the text and end up with a normal image of both. Is the latent space for "butterfly and snail" just well off the path between the two individually?
It's hard to imagine what is nearby in latent space and how text contributes, so I did really like the section adding words to the prompt 1-by-1.
Amazing explanations!! I absolutely love this. In 10 minutes it’s given me a huge boost in my intuition on diffusion, which I’ve been missing for years.
If the prompt is the compass, and represents a point in space, why walk there? Why not just go to that point in image space directly, what would be there? When does the random seed matter if you're aiming at the same point anyway, don't you end up there? Does the prompt vector not exist in the image manifold, or is there some local sampling done to pick images which are more represented in the training data?
So I’m not an expert, this post was just based on my understanding, but as I understand it: the prompt embedding space and the latent image space are different “spaces”, so there is no single “point” in the latent image space that represents a given prompt. There are regions that are more or less consistent with the prompt, and due to cross-attention between the text embedding vector and the latent image vector, it’s able to guide the diffusion process in a suitable direction.
So different seeds lead to slightly different end points, because you’re just moving closer to the “consistent region” at each step, but approaching from a different angle.
This is awesome. If you made a book or video-course that takes this level of high level explanation and translate it to the technical and then mathematical level, I would buy it in an heartbeat.
This is what I think is missing in most AI (broad sense) learning resources. They focus too much on the math that I miss the intuitive process behind it.
Hey, I made this, thanks for posting!
It’s purposefully high level and non-technical for a general audience - my theory was that most people who aren’t into tech/AI don’t care too much about training, or how the system got to be the way that it is.
But they do have some interest in how it actually operates once you’ve typed in a prompt.
Happy to answer any questions or take on board feedback
I think some of the visualizations would be much better if you used a pixel-space model instead of a latent diffusion model.
Right now we are only seeing the denoising process after it's been morphed by the latent decoder, which looks a lot less intuitive than actual pixel diffusion.
If you can't find a suitable pixel-space model, then you can just trivially generate a forward process and play it backwards.
Thanks that’s a great suggestion.
Loved the writeup!
Found the manual latent space exploration part really interesting.
Too many LLM/diffusion explanations fall in the proverbial “how to draw an owl” meme without giving a taste as to what’s going on.
It's quite clever and thoughtful. thanks for making it!
I enjoyed this a lot.
The interpolations between butterfly and snail were pretty horrifying. But something like Z-Image you could basically concatenate the text and end up with a normal image of both. Is the latent space for "butterfly and snail" just well off the path between the two individually?
It's hard to imagine what is nearby in latent space and how text contributes, so I did really like the section adding words to the prompt 1-by-1.
Oh I particularly loved that you made the prompts themselves interchangeable. Very well done!
Amazing explanations!! I absolutely love this. In 10 minutes it’s given me a huge boost in my intuition on diffusion, which I’ve been missing for years.
If the prompt is the compass, and represents a point in space, why walk there? Why not just go to that point in image space directly, what would be there? When does the random seed matter if you're aiming at the same point anyway, don't you end up there? Does the prompt vector not exist in the image manifold, or is there some local sampling done to pick images which are more represented in the training data?
So I’m not an expert, this post was just based on my understanding, but as I understand it: the prompt embedding space and the latent image space are different “spaces”, so there is no single “point” in the latent image space that represents a given prompt. There are regions that are more or less consistent with the prompt, and due to cross-attention between the text embedding vector and the latent image vector, it’s able to guide the diffusion process in a suitable direction.
So different seeds lead to slightly different end points, because you’re just moving closer to the “consistent region” at each step, but approaching from a different angle.
Scrolling through pics on mobile is difficult. Wanted to see all 29 steps but couldnt scroll it reliably.
Turning off the scroll mode worked very well for me on a mobile.