Creation of a computational space with model-free metasurface neural network
Mingfei Song, Chao Qian, Gaoang Wang, Cheng Li, Fang Yuan, Ali Momeni, Romain Fleury, Hongsheng Chen · 2026
Read Creation of a computational space with model-free metasurface neural network by Mingfei Song, Chao Qian, Gaoang Wang, Cheng Li, Fang Yuan, Ali Momeni, Romain Fleury, Hongsheng Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13439180.txt.
Description
Mingfei Song and colleagues build a computing space out of smart surfaces and radio waves, describing the system in 2026. Physical neural networks use real materials to compute, which promises speed and low energy use. Chaotic settings scramble signals, though, and exact maths models of them are hard to build. The team let the messy setting itself do the computing, spread across many tunable metasurface panels.
Training ran fully forward, using live radio feedback to estimate gradients instead of a model of the room. The space named handwritten letters and digits with high accuracy in tests. It also sharpened focusing through disordered scattering and located people in the room reliably. The authors frame the setup as a step toward machines that think through their physical surroundings.
Catalog record
| CREATOR | Mingfei Song, Chao Qian, Gaoang Wang, Cheng Li, Fang Yuan, Ali Momeni, Romain Fleury, Hongsheng Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60816 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |