G^3VLA: Geometric Inductive Bias for Vision-Language-Action Models ▲ 4 💬 1 2026-10-16 arXiv HF 原文 摘要 (无摘要)