GeoAlign: Beyond Semantics with State-Guided Spatial Alignment in VLA Models
返回详情
VLA / Vision-Language-Action 每日论文卡
设备