@misc{indiciae4748191f9b39,
  title = {Harnessing Image Question Dependence for Better VLM Test-time Reinforcement Learning},
  author = {Xinrui He and Ting-Wei Li and Junting Wang and Mengting Ai and Xinyu He and Hanghang Tong and Jingrui He},
  year = {2026},
  url = {https://arxiv.org/abs/2609.13296},
  note = {Source identifier: 2609.13296}
}