Vision Foundation Models Can Be Good Tokenizers for Latent Diffusion Models Paper • 2510.18457 • Published Oct 21, 2025 • 3