Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers Paper • 2607.19139 • Published 15 days ago • 74
VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis Paper • 2605.22570 • Published May 21 • 24
PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion Paper • 2605.23902 • Published May 22 • 47
Rethinking Cross-Layer Information Routing in Diffusion Transformers Paper • 2605.20708 • Published May 20 • 113