@article{whenllavameetsobjectstokencompositionfor,
  title = {When LLaVA Meets Objects: Token Composition for Vision-Language-Models},
  author = {Soumya Jahagirdar and Walid Bousselham and Anna Kukleva and Hilde Kuehne},
  year = {2026},
  eprint = {2602.04864},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2602.04864},
}