VideoAnchor: Reinforcing Subspace-Structured Visual Cues for Coherent Visual-Spatial Reasoning

@misc{wang2025videoanchorreinforcingsubspacestructuredvisual,
      title={VideoAnchor: Reinforcing Subspace-Structured Visual Cues for Coherent Visual-Spatial Reasoning},
      author={Zhaozhi Wang and Tong Zhang and Mingyue Guo and Yaowei Wang and Qixiang Ye},
      year={2025},
      eprint={2509.25151},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2509.25151},
}