UAOR: Uncertainty-aware Observation Reinjection for Vision-Language-Action Models figure
AlphaXiv 中文论文页面(可滚动查看)