@misc{indiciae6e6ba4306e05,
  title = {Zellige: Moldable Sequence Placement for Mixed Image-Video DiT Training},
  author = {Guangyu Xiang and Xueze Kang and Minwei Zhao and Yuxin Wang and Shaohuai Shi and Lin Zhang and Xiaowen Chu},
  year = {2026},
  url = {https://arxiv.org/abs/2608.01150},
  note = {Source identifier: 2608.01150}
}