LARY: A Latent Action Representation Yielding Benchmark for Generalizable Vision-to-Action Alignment

返回详情 VLA / Vision-Language-Action 每日论文卡