1 comment

[ 2.8 ms ] story [ 17.4 ms ] thread
summary by OP on reddit:

> Researchers experimentally discovered that image-generating AI Stable Diffusion v1 uses internal representations of 3D geometry - depth maps and object saliency maps - when generating an image. This ability emerged during the training phase of the AI, and was not programmed by people.

https://www.reddit.com/r/StableDiffusion/comments/15wvz2a/re...