Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models figure
AlphaXiv 中文概览(可滚动查看)