Embodied AI TopConf · ICLR2026

SP-VLA: A Joint Model Scheduling and Token Pruning Approach for VLA Model Acceleration

ICLR2026 / Vision-Language-Action Models

视觉语言动作感知加速/部署
新标签打开 AlphaXiv
SP-VLA: A Joint Model Scheduling and Token Pruning Approach for VLA Model Acceleration figure
AlphaXiv 中文论文页面（可滚动查看）