VLA-OS: Structuring and Dissecting Planning Representations and Paradigms in Vision-Language-Action Models figure
AlphaXiv 中文论文页面(可滚动查看)