Interpretable AI reveals hidden structure in nuclear masses.
4 affiliations
- 1NSF AI Institute for Artificial Intelligence and Fundamental Interactions, Cambridge, MA USA.
- 2Laboratory for Nuclear Science, MIT, Cambridge, MA USA.
- 3Theoretical Physics Department, CERN, Geneva, Switzerland.
- 4Physik-Institut, Universität Zürich, Zürich, Switzerland.
Abstract
Predicting nuclear masses with high precision remains a central challenge in nuclear physics, with implications ranging from stellar nucleosynthesis to determining the boundaries of the nuclear landscape. Approaches using artificial intelligence (AI) have recently achieved remarkable accuracy on this task, but their usefulness is limited by their black-box nature and by the difficulty of assessing their reliability when extrapolating beyond measured nuclei. Here, we present an interpretable AI model that matches the best existing predictive precision while revealing the physical structure underlying nuclear binding energies. The model learns internal representations that form a double helix, in which links between proton and neutron embeddings correspond to the most stable nuclei within each isotopic chain. Analysis of its learned features shows that the AI autonomously recovers established macroscopic components such as the liquid-drop model and rediscovers a long-forgotten physical principle-localized Jaffe factorization-that accounts for much of its performance gains. These findings demonstrate that interpretable AI can both reproduce and extend nuclear models, uncovering latent physical symmetries directly from data.
Related
- Bridging Ab Initio Symmetries and Global Nuclear Masses with Interpretable Neural Networks
- QCD sum rule analysis of hidden strange 1++ tetraquark masses and radial excitations
- Interpretable hybrid nuclear mass prediction based on term-by-term model discrepancies
- Effect of inelastic scalar dark matter in hidden U(1) scenario after the LZ nuclear recoil
- Identifying hidden nuclear material in orbit from its own neutron spectrum
BibTeX
@article{richardson2026interpretable,
title = {{Interpretable AI reveals hidden structure in nuclear masses.}},
author = {Kate A Richardson and Sokratis Trifinopoulos and Mike Williams},
journal = {Communications AI \& computing},
year = {2026},
month = oct,
doi = {10.1038/s44488-026-00017-9},
url = {https://europepmc.org/article/MED/42923159},
note = {CC BY},
}
RIS
TY - JOUR
TI - Interpretable AI reveals hidden structure in nuclear masses.
AU - Kate A Richardson
AU - Sokratis Trifinopoulos
AU - Mike Williams
PY - 2026
DA - 2026/10/21
JO - Communications AI & computing
DO - 10.1038/s44488-026-00017-9
UR - https://europepmc.org/article/MED/42923159
AB - Predicting nuclear masses with high precision remains a central challenge in nuclear physics, with implications ranging from stellar nucleosynthesis to determining the boundaries of the nuclear landscape. Approaches using artificial intelligence (AI) have recently achieved remarkable accuracy on this task, but their usefulness is limited by their black-box nature and by the difficulty of assessing their reliability when extrapolating beyond measured nuclei. Here, we present an interpretable AI model that matches the best existing predictive precision while revealing the physical structure underlying nuclear binding energies. The model learns internal representations that form a double helix, in which links between proton and neutron embeddings correspond to the most stable nuclei within each isotopic chain. Analysis of its learned features shows that the AI autonomously recovers established macroscopic components such as the liquid-drop model and rediscovers a long-forgotten physical principle-localized Jaffe factorization-that accounts for much of its performance gains. These findings demonstrate that interpretable AI can both reproduce and extend nuclear models, uncovering latent physical symmetries directly from data.
SN - 3091-292X
ER -