@misc{indiciae7a1ec022eea1,
  title = {Step-Level Visual Grounding Faithfulness Predicts Out-of-Distribution Generalization in Long-Horizon Vision-Language Models},
  author = {Md Ashikur Rahman and Md Arifur Rahman and Niamul Hassan Samin and Abdullah Ibne Hanif Arean and Juena Ahmed Noshin},
  year = {2026},
  url = {https://arxiv.org/abs/2603.06828},
  note = {Source identifier: 2603.06828}
}