TY  - RPRT
TI  - Learning from Runtime Feedback through Failure-Bank Self-Evolution for Vision-Language-Action Models
AU  - Mingyue Cui
AU  - Zheyuan Liu
AU  - Yihan Zhu
AU  - Zheyuan Zhang
AU  - Meng Jiang
PY  - 2026
UR  - https://arxiv.org/abs/2609.39820
ID  - 2609.39820
ER  -