LiFT: Looping Shared DiT Core Cuts Parameters 60%, Improves FID

The LiFT paper proposes looping a shared Diffusion Transformer core to scale compute, cutting parameters by 60% while improving FID by 3.34 over dense baselines.

2026-10-21 ~ 2026-10-21 · 2 related posts

• LiFT loops a shared DiT core to beat dense baseline by 3.34 FID with 60% fewer params — burny_tech · 2026-10-21

1 near-duplicate retellings: burny_tech