PaperRepresentation Conditioning for Generation
Representation Conditioning for Generation
An academic paper that proposes the RECON representation conditioning method, aimed at accelerating the training convergence of diffusion Transformers.
Timeline (last 90 days)
Oct 15
The paper "Representation Conditioning for Generation" proposes the RECON method, which, without adding any parameters, speeds up diffusion Transformer training convergence to 63x that of SiT and 25x that of REPA.
Unverified50%