SDAR-VL: Stable and Efficient Block-wise Diffusion for Vision-Language Understanding

Add code
Dec 16, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: