S$^2$-VLA: State-Space Guided Vision-Language-Action Models for Long-Horizon Manipulation figure
AlphaXiv 中文概览(可滚动查看)