What Do VLAs Actually Learn through In-Context Failure Conditioning?

Jiajun Liu, Jieming Li, Zi Zhuang, Hang Yu, Qingli Chen, Liu Cao, Yingxi Lu, Ruoqu Chen, Yuhang Cao, Chenyu Zhang, Yankai Lin, Mengdi Xu

Published in CVPR 2026 Workshop on 3D-LLM/VLA, 2026