TY  - RPRT
TI  - Your Vision-Language-Action Model Already Has Attention Heads For Path Deviation Detection
AU  - Jaehwan Jeong
AU  - Evelyn Zhu
AU  - Jinying Lin
AU  - Emmanuel Jaimes
AU  - Tuan-Anh Vu
AU  - Jungseock Joo
AU  - Sangpil Kim
AU  - M. Khalid Jawed
PY  - 2026
UR  - https://arxiv.org/abs/2603.13782
ID  - 2603.13782
ER  -