Both Semantics and Reconstruction Matter: Making Representation Encoders Ready for Text-to-Image Generation and Editing

Open paper

Benchmark	Model	Rank	Results
text-to-image-generation-on-dpg-bench	PS-VAE (32c)	–	DPG-Bench Score: 84.25
text-to-image-generation-on-geneval	PS-VAE (96c, SigLIP2-so400m/14)	–	Overall: 0.7714