Paper link: <a href="https://arxiv.org/pdf/2209.11178.pdf" rel="nofollow">https://arxiv.org/pdf/2209.11178.pdf</a><p>Lots of interesting things I'd want to try here. Eg, the formulation of thinking of your probability distribution as a charge distribution opens up the possibility of adding negative charges. So you could have a regular positive data examples, but also some negative examples that the generative model tries extra hard to avoid.<p>I'd also be interested to see what happens when you "add the fields" for two different distributions over the same space. Seems like it would be interestingly different than regular superposition, and I'd really like to see what happens when you add together a dog image field to a cat image field.<p>I guess the current formulation where you only get the direction of the electric field and not the magnitude would cause problems there. Seems like an improvement might be to predict the actual potential instead of the field and then take the gradient. That would also guarantee that the field is curl-free, which it should be, but it doesn't seem to be enforced in the current paper. I think they don't try this because it's hard to get NNs to produce numbers across a wide range of orders of magnitude. Maybe separately predicting the normalized direction and something like the log of the magnitude would help here.
All the novelty appears to boil down to "this is a new sampler for diffusion-style models".<p>Apparently, they are inspired by a different physical process. But that doesn't change the fact that they start with noise and then iteratively solve a differential equation to get to its end point. 2nd-order samplers like Heun already massively increased sampling speed over the original diffusion models, but sadly the article doesn't compare to that but merely to the original 1st-order samplers. So it might be that this one doesn't even create a speedup in practice.<p>Does anyone else have more info on how this is different from "traditional" numerical differential-equation-based de-noising?
> PFGMs constitute an exciting foundation for new avenues of research, especially given that they are 10-20 times faster than Diffusion Models on image generation tasks, with comparable performance.<p>hard to imagine anyone sticking with diffusion models if this is completely true