Data and code for "Multi-nuclide identification in HPGe gamma-ray spectra using physics-informed peak features and nuclide-wise neural networks"

It includes FLUKA-simulated and measured HPGe event pools and spectra, four-, ten-, and nineteen-nuclide study results, trained PyTorch models, configurations, prediction tables, figure and table source data, and reproducibility code. The released files support recalculation of the reported results, model re-evaluation, and regeneration of training and evaluation datasets.

README files, data dictionaries, and manifests document the file structure, variables, units, provenance, and checksums.Data files and model artifacts are released under CC BY 4.0. The three code packages are released under the MIT License, as specified in their respective LICENSE files.