📡 Embodied AI Radar具身智能雷达 Star2

Most of this project is curated by Claude. Drop a star to cheer Claude on!本项目大量由 Claude 整理维护,点个 Star 给克劳德大小姐鼓励一下!

★ Feed a star投喂 Star

Topics主题 › VLA & Robot Foundation ModelsVLA 与机器人基础模型

Paper论文 2026 CoRL 2026 Spotlight new新 2026-11-02

When Vision Overrides Language: Evaluating and Mitigating Counterfactual Failures in VLAs

CoRL 2026 Spotlight. VLAs can ignore language and follow vision shortcuts from dataset bias. Introduces LIBERO-CF, a counterfactual benchmark for language following.CoRL 2026 Spotlight。VLA 可能忽略语言、沿用数据偏差形成的视觉捷径。论文提出反事实基准 LIBERO-CF,用于评测语言遵循能力。

BibTeX
@misc{fang2026when,
  title = {{When Vision Overrides Language: Evaluating and Mitigating Counterfactual Failures in VLAs}},
  author = {Yu Fang and Yuchun Feng and Dong Jing and Jiaqi Liu and Yue Yang and Zhenyu Wei and Daniel Szafir and Mingyu Ding},
  year = {2026},
  eprint = {2602.17659},
  archivePrefix = {arXiv},
  primaryClass = {cs.CV},
  url = {https://arxiv.org/abs/2602.17659},
}
VLA & Robot Foundation ModelsVLA 与机器人基础模型counterfactualbenchmarklanguage-following

Authors (from arXiv): 作者(来自 arXiv):Yu Fang, Yuchun Feng, Dong Jing, Jiaqi Liu, Yue Yang, Zhenyu Wei, Daniel Szafir, Mingyu Ding

Code license代码许可证 (detected by GitHub): (GitHub 检测):MIT

Similar entries相似条目

All VLA & Robot Foundation Models entries →查看全部VLA 与机器人基础模型条目 →