Can Vision-Language-Action Models Learn from Real-World Data Continually without Forgetting?

@misc{zhu2026visionlanguageaction,
  title = {Can Vision-Language-Action Models Learn from Real-World Data Continually without Forgetting?},
  author = {Jiarun Zhu and Yijun Hong and Xiao-Quan Sun and Ze-Tian Xu and Qi-Jun He and Haijier Chen and Zhiyong Wang and Mingqi Yuan and Wenjun Zeng and Jia-Yu Chen},
  year = {2026},
  eprint = {2605.26820},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.26820},
}