@misc{indiciae810244cc50ad,
  title = {EVATok: Adaptive Length Video Tokenization for Efficient Visual Autoregressive Generation},
  author = {Tianwei Xiong and Jun Hao Liew and Zilong Huang and Zhijie Lin and Jiashi Feng and Xihui Liu},
  year = {2026},
  url = {https://arxiv.org/abs/2603.12267},
  note = {Source identifier: 2603.12267}
}