Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment figure
AlphaXiv 中文概览(可滚动查看)