@misc{indiciaecb927eb3ce5f,
  title = {VANE: Reliable Test-Time Training for Vision-Language-Action Models via Future Visual Representation Prediction},
  author = {Hongjin Ji and Guoyang Xia and Luoyang Sun and Fangxiang Feng and Lei Ren},
  year = {2026},
  url = {https://arxiv.org/abs/2608.09448},
  note = {Source identifier: 2608.09448}
}