@misc{indiciaebdb1a76a5498,
  title = {CatchUp: Depth-Aware Re-Admission of Visual Tokens for Efficient Large Vision-Language Models},
  author = {Eunju Park and Uiwon Hwang},
  year = {2026},
  url = {https://arxiv.org/abs/2610.23324},
  note = {Source identifier: 2610.23324}
}