A paper on 1-dimensinal tokenization and text-to-image generation is accepted at ICCV 2025.