@misc{indiciae9a56847d1121,
  title = {Pretrained Vision-Language-Action Models are Surprisingly Resistant to Forgetting in Continual Learning},
  author = {Huihan Liu and Changyeon Kim and Bo Liu and Minghuan Liu and Yuke Zhu},
  year = {2026},
  url = {https://arxiv.org/abs/2603.03818},
  note = {Source identifier: 2603.03818}
}