1 paper
Siya Yang, Nanxiang Jiang, Zhaoxin Fan +1
The rapid progress of visual autoregressive (VAR) models has unlocked a transformative frontier for high-fidelity text-to-image synthesis, while heightening concerns over the safet…