Embodied AI TopConf · NeuIPS2025

VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching

NeuIPS2025 / Accelerating and Deploying

VLA感知加速/部署
新标签打开 AlphaXiv新标签打开 Project
VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching figure
AlphaXiv 中文论文页面（可滚动查看）