Patrick Esser

Rating

Domains

Scouting report

Co-authored VQGAN — the taming-transformers work that made high-resolution generation tractable — and then latent diffusion and Stable Diffusion. Later led the rectified-flow architecture behind Stable Diffusion 3 and moved to Black Forest Labs for the FLUX models, which were also released with open weights.

Curator’s note

Two labs, three architectures, and the weights kept coming out.

Portrait

Painted from a photograph of Patrick Esser by Patrick Esser, supplied by the subject.

Sources

• 01Taming Transformers for High-Resolution Image Synthesis (2020) arxiv.org
• 02Scaling Rectified Flow Transformers (2024) arxiv.org