LiFT: Loop Flow Transformers
LiFT loops a shared Diffusion Transformer and lowers ImageNet FID while cutting parameters and FLOPs.
Stories per day · 30 days
2 stories, newest first · one card per story
Stories tagged #lift
Pretraining Latent Information Feedback Transformers with Teacher Supervision
LIFT pretrains Transformers to recycle latent state and beats token-matched baselines from 135M to 1B.