arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of California, Berkeley(加州大学伯克利分校)

2026-08-05 至 2026-08-05 共收录 4
2605.24011 2026-08-05 cs.CV cs.AI 版本更新

ActQuant: Sub-4-bit Action-Guided Quantization for Vision-Language-Action Models

ActQuant: 面向视觉-语言-动作模型的亚4比特动作引导量化

Arash Akbari, Arman Akbari, Masih Eskandar, Qitao Tan, Yixiao Chen, Jingwu Luo, Bertha Pangaribuan, Liyun Zhang, Jennifer Dy, Geng Yuan, Xue Lin, Gaowen Liu, Stratis Ioannidis, Yanzhi Wang

机构 * Northeastern University(东北大学) University of California, Berkeley(加州大学伯克利分校)

AI总结 提出ActQuant框架,通过动作引导的混合精度后训练量化,在亚4比特权重量化下保持VLA模型性能,并引入OmniModel.cpp实现高效部署。

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.19209 2026-08-05 cs.RO cs.MA 版本更新

Graph Neural Planning and Predictive Control for Multi-Robot Communication-Constrained Unlabeled Motion Planning

基于图神经网络的多机器人通信受限的无标签运动规划与预测控制

Manohari Goarin, Yang Zhou, Giuseppe Loianno

机构 * New York University(纽约大学) University of California Berkeley(加州大学伯克利分校)

AI总结 本文提出一种分层框架,结合图注意力规划器和分布式非线性模型预测控制器,以解决多机器人在通信受限环境下同时分配目标和生成安全轨迹的问题,通过图神经网络方法实现可扩展的去中心化解决方案。

Comments 8 pages, 6 figures, Accepted at the IEEE International Conference on Robotics and Automation (ICRA) 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.20821 2026-08-05 cs.DC cs.AI cs.LG

Compass: Optimizing Compound AI Workflows for Dynamic Adaptation

Compass: 为动态适应优化复合AI工作流

Milos Gravara, Juan Luis Herrera, Stefan Nastic

机构 * University of California, Berkeley(加州大学伯克利分校) ETH Zurich(苏黎世联邦理工学院)

AI总结 本文提出Compass框架,通过离线优化和在线适应动态切换复合AI工作流的配置,提升准确率、延迟和成本的平衡能力。

Comments 10 pages, 7 figures; accepted at the 26th IEEE International Symposium on Cluster, Cloud, and Internet Computing (CCGrid 2026)

Journal ref Proceedings of the 2026 IEEE 26th International Symposium on Cluster, Cloud and Internet Computing (CCGrid), pp. 84-93, 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2502.11140 2026-08-05 cs.SE cs.AI cs.CL cs.HC 版本更新

Automated Visualization Code Synthesis via Multi-Path Reasoning and Feedback-Driven Optimization

通过多路径推理和反馈驱动优化实现自动化可视化代码合成

Wonduk Seo, Daye Kang, Hyunjin An, Taehan Kim, Soohyuk Cho, Seungyong Lee, Minhyeong Yu, Jian Park, Yi Bu, Seunghyun Lee

机构 * AI Research, Enhans, Seoul, South Korea Innovation \& Technology, KAIST, Daejeon, South Korea Department of Computer Science, University of California, Berkeley, CA, United States Department of Electrical

AI总结 VisPath通过多路径推理和反馈驱动优化,提升自动化可视化代码生成的可靠性与准确性。

Comments Accepted by International Conference on Pattern Recognization (ICPR 2026)

Journal ref Pattern Recognition. ICPR 2026. Lecture Notes in Computer Science, vol 16812. Springer, Cham

详情

展开后加载摘要…

URL PDF HTML 收藏