@article{feat2govisualfeaturegroundedvalueestimat,
  title = {Feat2Go: Visual Feature-Grounded Value Estimation for Embodied Reinforcement Learning},
  author = {Junyang Shu and Zhiwei Lin and Bingqing Wei and Yongtao Wang},
  year = {2026},
  eprint = {2605.30795},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.30795},
}