@misc{indiciae1703d2689452,
  title = {BARD: Bridging AutoRegressive and Diffusion Vision-Language Models Via Highly Efficient Progressive Block Merging and Stage-Wise Distillation},
  author = {Baoyou Chen and Hanchen Xia and Peng Tu and Haojun Shi and Liwei Zhang and Yuxuan Yao and Weihao Yuan and Siyu Zhu},
  year = {2026},
  url = {https://arxiv.org/abs/2604.16514},
  note = {Source identifier: 2604.16514}
}