arXiv preprint 2025 / 令牌优化

Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models

作者：Xudong Tan, Yaoxin Yang, Peng Ye, Jialin Zheng, Bizhe Bai, Xinyi Wang, Jia Hao, Tao Chen · 单位：Fudan University Shanghai AI Laboratory, The Chinese University of Hong Kong Zhangjiang Laboratory · 会议/期刊：arXiv preprint 2025 · 方向：令牌优化 · 发布日期：2025-05-27

视觉语言动作令牌优化高效推理模型压缩
新标签打开 AlphaXiv
Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models figure
AlphaXiv 中文概览（可滚动查看）