Speridlabs' Iris-3B Generates Megapixel Images Without a VAE
Speridlabs released a 3B pixel-space diffusion transformer that skips the VAE entirely, generating 1024px images and doubling as a depth and restoration backbone.
Source
AlphaSignal
Published
Author
AlphaSignal Newsroom
Read
1 min read
Speridlabs released a 3B pixel-space diffusion transformer that skips the VAE entirely, generating 1024px images and doubling as a depth and restoration backbone.
Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.
