SwiftVLA: Unlocking Spatiotemporal Dynamics for Lightweight VLA Models at Minimal Overhead
返回详情
VLA / Vision-Language-Action 每日论文卡