In situ local learning of dynamic network materials
Abstract
Functional materials realize functionalities through pre-designed structures, whereas physical neural networks are trained by prescribing input-output behavior. Unifying these views would allow materials to acquire functions directly from dynamical tasks. Here we introduce an in situ local learning framework for dynamic network materials. A forward drive and a time-reversed adjoint error drive are applied to the same mechanical network, enabling the material to compute gradients of a time-domain loss through its own dynamics. The resulting update rules are local, compatible with the principle of physical learning. Each bond or node requires only co-located forward and adjoint fields. The method trains spring constants, nodal masses, bond damping coefficients and nodal damping coefficients, and programs dynamical functions including broadband wave cloaking, evanescent-wave imaging, passive transient output enhancement and vowel classification. These examples involve distinct physical mechanisms, such as wave scattering, near-field information transfer, dissipative non-normal dynamics and machine learning, but are obtained by simply changing the loss function, rather than leveraging expert knowledge. Therefore, dynamic network materials can act both as programmable matter and as physical neural networks, providing a route to material systems that learn temporal responses in situ.
Related
- Accelerating dynamic simulations of photoexcited materials and their evolution by electron-informed machine learning
- Contact mechanics and friction of soft materials: an apparatus combining multi-axes dynamical actuation/measurement and in situ/in operando visualisation
- Small scale mechanical properties of polycrystalline materials: in situ diffraction studies
- Symmetry-Preserving Local Representation for Short-Range Interaction in Complex Materials
- Dynamic moiré-like band modulation in Dirac materials via multi-beam optical interference
BibTeX
@misc{li2026situ,
title = {{In situ local learning of dynamic network materials}},
author = {Shuaifeng Li and Xiaoming Mao},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.05977},
eprint = {2609.05977},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.05977},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - In situ local learning of dynamic network materials
AU - Shuaifeng Li
AU - Xiaoming Mao
PY - 2026
DA - 2026/09/05
PB - arXiv
DO - 10.48550/arxiv.2609.05977
UR - https://arxiv.org/abs/2609.05977
AB - Functional materials realize functionalities through pre-designed structures, whereas physical neural networks are trained by prescribing input-output behavior. Unifying these views would allow materials to acquire functions directly from dynamical tasks. Here we introduce an in situ local learning framework for dynamic network materials. A forward drive and a time-reversed adjoint error drive are applied to the same mechanical network, enabling the material to compute gradients of a time-domain loss through its own dynamics. The resulting update rules are local, compatible with the principle of physical learning. Each bond or node requires only co-located forward and adjoint fields. The method trains spring constants, nodal masses, bond damping coefficients and nodal damping coefficients, and programs dynamical functions including broadband wave cloaking, evanescent-wave imaging, passive transient output enhancement and vowel classification. These examples involve distinct physical mechanisms, such as wave scattering, near-field information transfer, dissipative non-normal dynamics and machine learning, but are obtained by simply changing the loss function, rather than leveraging expert knowledge. Therefore, dynamic network materials can act both as programmable matter and as physical neural networks, providing a route to material systems that learn temporal responses in situ.
ER -