biostudies-literatureYu ZNatural Science Foundation of Jiangsu Province (Jiangsu Provincial Natural Science Foundation)Jiangsu Planned Projects for Postdoctoral Research FundsChina Postdoctoral Science FoundationNational Natural Science Foundation of China (National Science Foundation of China)127https://www.ebi.ac.uk/biostudies/studies/S-EPMC12946256biostudies-literatureUnknown15(1)Optical computing holds significant promise across diverse applications due to its low latency, power efficiency, and multidimensional processing capabilities. However, current diffraction neural networks (DNNs) generally lack reconfigurability, limiting the scalability of the optical computing systems. Inspired by movable-type printing technology, here we propose a movable-type coding metasurface to enable multiple functionalities such as electromagnetic (EM) computing, holography, and sensing. By cascading multiple layers of the proposed metasurfaces, we further develop a movable-type reconfigurable DNN (MT-RDNN). It can be seamlessly adapted from handwritten digit to letter classification tasks by replacing the meta-atoms in the last hidden metasurface layer. Moreover, a single-layer moLight, science & applicationsMultifunctional movable-type coding metasurface enabling reconfigurable diffractive neural networks.PMC129462562023ZB318No. BK20210209 and BK20230820No. 2024M75041862288101, 62301149Li XMa QChen LCai ZGu ZCui TJSu JZhang LYu ZYu XQin SYou JWfalseMultifunctional movable-type coding metasurface enabling reconfigurable diffractive neural networks.Optical computing holds significant promise across diverse applications due to its low latency, power efficiency, and multidimensional processing capabilities. However, current diffraction neural networks (DNNs) generally lack reconfigurability, limiting the scalability of the optical computing systems. Inspired by movable-type printing technology, here we propose a movable-type coding metasurface to enable multiple functionalities such as electromagnetic (EM) computing, holography, and sensing. By cascading multiple layers of the proposed metasurfaces, we further develop a movable-type reconfigurable DNN (MT-RDNN). It can be seamlessly adapted from handwritten digit to letter classification tasks by replacing the meta-atoms in the last hidden metasurface layer. Moreover, a single-layer mo2026-01-01T00:00:00Z2026 Feb2026-07-16T22:16:53.068Z2026-07-12T03:08:44.132ZS-EPMC129462564174853610.1038/s41377-026-02216-6