@misc{indiciae18aa2a0a333d,
  title = {Saliency-R1: Enforcing Interpretable and Faithful Vision-language Reasoning via Saliency-map Alignment Reward},
  author = {Shizhan Gong and Minda Hu and Qiyuan Zhang and Chen Ma and Qi Dou},
  year = {2026},
  url = {https://arxiv.org/abs/2604.04500},
  note = {Source identifier: 2604.04500}
}