@misc{indiciae154590181d67,
  title = {Prioritizing Image-Related Tokens Enhances Vision-Language Pre-Training},
  author = {Yangyi Chen and Hao Peng and Tong Zhang and Heng Ji},
  year = {2025},
  url = {https://arxiv.org/abs/2505.08971},
  note = {Source identifier: 2505.08971}
}