@misc{indiciae1a3b03068487,
  title = {DAVET: Denoising-Aware Visual Evidence Trajectory Allocation for Diffusion Vision-Language Models},
  author = {Yongkang Zhou and Xiang Xia and Cheng Yan and Fan Xu and Wuyang Zhang},
  year = {2026},
  url = {https://arxiv.org/abs/2608.01821},
  note = {Source identifier: 2608.01821}
}