Learning materials spectra across modalities and scales with structure-conditioned neural fields
Abstract
Spectroscopic observables encode rich information about material behavior across diverse, often high-dimensional domains of energy, momentum, and other experimental coordinates, providing critical insights into the physical properties underlying material functionalities and applications. However, learning these responses directly from atomic structures remains challenging due to the complexity of both structural configurations and spectroscopic observation spaces. Here, we develop a structure-conditioned neural-field framework that learns continuous spectroscopic response functions by coupling an equivariant structural encoder with a coordinate-conditioned network. The framework accurately predicts spectroscopic observables across diverse observation domains, ranging from one-dimensional densities of states to multidimensional scattering functions. We show that fine-tuning with limited spectral supervision captures strain- and composition-dependent spectral changes, with validated transfer to alloy configurations containing hundreds of atoms and computational scalability demonstrated for configurations exceeding 104 atoms. Our work establishes a versatile and computationally scalable approach to learning structure-spectroscopy relationships, providing a foundation for spectroscopic analysis and future materials characterization, inverse problems, and experimental design.
Related
- Wyckoff-Resolved Oxidation-State Atlas and Anion-Conditioned Priors for Materials Discovery
- Thermally Conditioned Time-Series Forecasting of Grain Growth in Polycrystalline Materials
- PXtal: Learning to Align Powder X-Ray Diffraction and Crystal Structures under Information Asymmetry across Modalities
- Comprehensive reconstruction of collider events with hypergraph representation learning and graph-conditioned diffusion
- Influence of External Electric Field on B6C6N6: Focusing on Structure, Electronic Structure, Spectra and Reactivity
BibTeX
@misc{zhang2026learning,
title = {{Learning materials spectra across modalities and scales with structure-conditioned neural fields}},
author = {Tianyu Zhang and Yuan Ni and Roy Liu and Garrett Granroth and Daniel M. Pajerowski and Joshua J. Turner and Yongqiang Cheng and Zhantao Chen},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01417},
eprint = {2611.01417},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01417},
}
RIS
TY - UNPB
TI - Learning materials spectra across modalities and scales with structure-conditioned neural fields
AU - Tianyu Zhang
AU - Yuan Ni
AU - Roy Liu
AU - Garrett Granroth
AU - Daniel M. Pajerowski
AU - Joshua J. Turner
AU - Yongqiang Cheng
AU - Zhantao Chen
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01417
UR - https://arxiv.org/abs/2611.01417
AB - Spectroscopic observables encode rich information about material behavior across diverse, often high-dimensional domains of energy, momentum, and other experimental coordinates, providing critical insights into the physical properties underlying material functionalities and applications. However, learning these responses directly from atomic structures remains challenging due to the complexity of both structural configurations and spectroscopic observation spaces. Here, we develop a structure-conditioned neural-field framework that learns continuous spectroscopic response functions by coupling an equivariant structural encoder with a coordinate-conditioned network. The framework accurately predicts spectroscopic observables across diverse observation domains, ranging from one-dimensional densities of states to multidimensional scattering functions. We show that fine-tuning with limited spectral supervision captures strain- and composition-dependent spectral changes, with validated transfer to alloy configurations containing hundreds of atoms and computational scalability demonstrated for configurations exceeding 10^4 atoms. Our work establishes a versatile and computationally scalable approach to learning structure-spectroscopy relationships, providing a foundation for spectroscopic analysis and future materials characterization, inverse problems, and experimental design.
ER -