@misc{indiciae5b8430427a8e,
  title = {GRACE: Boosting Video MLLMs with Grounded Action-Centric Evidence for Viewer Sentiment Prediction},
  author = {Ruoxuan Yang and Tieyuan Chen and Xiaofeng Huang and Haibing Yin and Jun Wang and Xiping Chen and Jun Yin and Xuesong Gao and Weiyao Lin},
  year = {2026},
  url = {https://arxiv.org/abs/2606.16198},
  note = {Source identifier: 2606.16198}
}