@misc{indiciae0e9968cfedb5,
  title = {What DINO saw: ALiBi positional encoding reduces positional bias in Vision Transformers},
  author = {Moritz Pawlowsky and Antonis Vamvakeros and Alexander Weiss and Anja Bielefeld and Samuel J. Cooper and Ronan Docherty},
  year = {2026},
  url = {https://arxiv.org/abs/2603.16840},
  note = {Source identifier: 2603.16840}
}