作者位次 #2/8ACL 2026Poster· Natural Language Processing
Stable Language Guidance for Vision–Language–Action Models
Zhihao Zhan · Yuhao Chen · Jiaying Zhou · Qinhan Lyu · Hao Liu · Keze Wang · Liang Lin · Guangrun Wang
摘要
Vision-Language-Action (VLA) models have demonstrated impressive capabilities in generalized robotic control; however, they remain notoriously brittle to linguistic perturbations. We identify a critical "modality collapse” phenomenon where strong visual priors overwhelm sparse linguistic signals,...