Embodied AI TopConf · NeuIPS2025

EfficientVLA: Training-Free Acceleration and Compression for Vision-Language-Action Models

NeuIPS2025 / Accelerating and Deploying

VLA感知加速/部署
新标签打开 AlphaXiv
EfficientVLA: Training-Free Acceleration and Compression for Vision-Language-Action Models figure
AlphaXiv 中文论文页面（可滚动查看）