@misc{indiciae337e5f1c5ede,
  title = {Parameter Efficient Multimodal Instruction Tuning for Romanian Vision Language Models},
  author = {George-Andrei Dima and Răzvan-Alexandru Smădu and Dumitru-Clementin Cercel},
  year = {2026},
  url = {https://arxiv.org/abs/2512.14926},
  note = {Source identifier: 2512.14926}
}