DriveMA: Rethinking Language Interfaces in Driving VLAs with One-Step Meta-Actions
返回详情
VLA / Vision-Language-Action 每日论文卡