If it's trained on absolute player coordinates then it would likely just morph i... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		WithinReason on Aug 28, 2024 \| parent \| context \| favorite \| on: Diffusion models are real-time game engines If it's trained on absolute player coordinates then it would likely just morph into the known map at those coordinates.

nine_k on Aug 28, 2024 [–]

But it's trained on the actual screen pixel data, AFAICT. It's literally a visual imagination model, not gameplay / geometry imagination model. They had to make special provisions to the pixel data on the HUD which by its nature different than the pictures of a 3D world.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact