@misc{indiciae79e989aeea58,
  title = {PairUni: Pairwise Training for Unified Multimodal Language Models},
  author = {Jiani Zheng and Zhiyang Teng and Kunpeng Qiu and Xiangtai Li and Anran Wang and Yu Tian and Ye Tian and Haochen Wang and Zhuochen Wang},
  year = {2026},
  url = {https://arxiv.org/abs/2510.25682},
  note = {Source identifier: 2510.25682}
}