SWIFT: Self-guided Weak-to-strong dIFfusion models via Target extrapolation Paper • 2610.25844 • Published 2 days ago • 21
Super-Resolution in The Right Latent Space: A Frozen Vision-Foundation Substrate Paper • 2610.04781 • Published 13 days ago • 26
Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers Paper • 2607.19139 • Published Jul 21 • 76
Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps Paper • 2605.16928 • Published May 16 • 92