arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

Beihang University(北京航空航天大学)

2026-06-08 至 2026-06-08 共收录 2
2605.25413 2026-06-08 cs.LG cs.AI cs.NA math.NA 版本更新

Autoregression-Free Neural Operators for Time-Dependent PDEs

无自回归的神经算子用于时间相关偏微分方程

Jiaquan Zhang, Caiyan Qin, Haoyu Bian, Libin Cai, Yi Lu, Chaoning Zhang, Wei Dong, Yuanfang Guo, Yang Yang, Heng Tao Shen

机构 * School of Computer Science and Engineering, University of Electronic Science and Technology of China(电子科技大学计算机科学与工程学院) School of Robotics and Advanced Manufacture, Harbin Institute of Technology(哈尔滨工业大学机器人与先进制造学院) School of Mathematical Sciences, Capital Normal University(首都师范大学数学学院) College of Information and Control Engineering, Xi’an University of Architecture and Technology(西安建筑科技大学信息与控制工程学院) Laboratory of Intelligent Recognition and Image Processing, School of Computer Science and Engineering, Beihang University(北京航空航天大学智能识别与图像处理实验室) School of Computer Science and Technology, Tongji University(同济大学计算机科学与技术学院)

AI总结 提出AFNO,通过将PDE时间演化映射到潜空间并利用流匹配学习连续时间向量场,避免自回归展开,实现长期稳定预测。

Comments 23 pages, 18 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.13042 2026-06-08 cs.LG cs.AR 版本更新

OpenACMv2: An Accuracy-Constrained Co-Optimization Framework for Approximate DCiM

OpenACMv2:面向近似数字存内计算的精度约束协同优化框架

Yiqi Zhou, Yue Yuan, Yikai Wang, Bohao Liu, Qinxin Mei, Zhuohua Liu, Shan Shen, Wei Xing, Daying Sun, Li Li, Guozhu Liu

机构 * Nanjing University of Science and Technology, China(南京理工大学) Shenzhen University, China(深圳大学) Beihang University, China(北京航空航天大学) University of Sheffield, UK(谢菲尔德大学) The 58th Research Institute of China Electronics Technology Group Corporation, China(中国电子科技集团第五十八研究所)

AI总结 提出OpenACMv2框架,通过两级优化(架构搜索和晶体管尺寸调整)实现精度约束下数字存内计算的功耗-性能-面积权衡,实验显示功耗延迟积降低50%以上。

Comments Accepted by DAC2026. Camera-ready version

详情

展开后加载摘要…

URL PDF HTML 收藏