1 paper
Zigang Geng, Yibing Wang, Yeyao Ma +10
Numerous efforts have been made to extend the ``next token prediction'' paradigm to visual contents, aiming to create a unified approach for both image generation and understanding…