Created on July 07, 2026
2026
LaViDa-R1, our reasoning model for unified multimodal diffusion, appears at ICML 2026.