Probing the many-body energy landscape of a soft glass with graph neural networks
Abstract
Many-body interactions govern the complex behavior of many amorphous materials, from metallic glasses to biological tissues, yet are often replaced by pairwise additive frameworks for computational efficiency. Here, we use classical density functional theory (DFT) to study a model soft glass of solvent-free polymer-grafted nanoparticles (PGNs), where the absence of solvent forces grafted chains to uniformly fill the interstitial space, generating strong angular-dependent many-body interactions between the cores. We show that NequIP, an equivariant message-passing graph neural network (GNN), predicts the equilibrium states with DFT-level accuracy, despite being trained only on high-energy, out-of-equilibrium configurations. Systematic analysis of GNN hyperparameters offers physical insights into the range, anisotropy, and effective body order of interactions. GNN-driven Monte Carlo simulations recover equilibrium structures in agreement with experiments and reveal locally favored icosahedral-like structures. Our results position machine-learned potentials as a tool for probing many-body glass physics beyond atomistic settings.
Related
- Probabilistic Energy-Based Neural Networks for Uncertainty-Aware Collider Data Analyses
- Data-Driven Discovery of a Hyperelastic Strain-Energy Function for Veins Using Constitutive Artificial Neural Networks
- Learning the self-energy through equivariant neural networks
- Physics-Informed Neural Networks to Infer the Perpendicular Energy Conductivity in the Scrape-Off Layer of Stellarator Devices
- Physics-augmented neural networks for predicting nonlinear vibration energy harvesters
BibTeX
@misc{ghomsheh2026probing,
title = {{Probing the many-body energy landscape of a soft glass with graph neural networks}},
author = {Mehryar Jannesari Ghomsheh and Donald L. Koch and Sarah Hormozi},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2605.30822},
eprint = {2605.30822},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.30822},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Probing the many-body energy landscape of a soft glass with graph neural networks
AU - Mehryar Jannesari Ghomsheh
AU - Donald L. Koch
AU - Sarah Hormozi
PY - 2026
DA - 2026/09/21
PB - arXiv
DO - 10.48550/arxiv.2605.30822
UR - https://arxiv.org/abs/2605.30822
AB - Many-body interactions govern the complex behavior of many amorphous materials, from metallic glasses to biological tissues, yet are often replaced by pairwise additive frameworks for computational efficiency. Here, we use classical density functional theory (DFT) to study a model soft glass of solvent-free polymer-grafted nanoparticles (PGNs), where the absence of solvent forces grafted chains to uniformly fill the interstitial space, generating strong angular-dependent many-body interactions between the cores. We show that NequIP, an equivariant message-passing graph neural network (GNN), predicts the equilibrium states with DFT-level accuracy, despite being trained only on high-energy, out-of-equilibrium configurations. Systematic analysis of GNN hyperparameters offers physical insights into the range, anisotropy, and effective body order of interactions. GNN-driven Monte Carlo simulations recover equilibrium structures in agreement with experiments and reveal locally favored icosahedral-like structures. Our results position machine-learned potentials as a tool for probing many-body glass physics beyond atomistic settings.
ER -