@misc{indiciaee55ae4137e93,
  title = {Training-Free Self-Correction for Multimodal Masked Diffusion Models},
  author = {Yidong Ouyang and Panwen Hu and Zhengyan Wan and Zhe Wang and Liyan Xie and Dmitriy Bespalov and Ying Nian Wu and Guang Cheng and Hongyuan Zha and Qiang Sun},
  year = {2026},
  url = {https://arxiv.org/abs/2602.02927},
  note = {Source identifier: 2602.02927}
}