Physical-Admissibility Projection and Split-Conformal Calibration for Burgers Neural Operators: Reproducibility Package

This record contains the numerical datasets, source code, trained model parameters, predictions, validation records, and reproducibility materials associated with the study “Physical-Admissibility Projection and Split-Conformal Calibration for Burgers Neural Operators.” The study combines mathematical analysis with computational experiments for the periodic viscous Burgers equation, including physical-admissibility projection, split-conformal calibration, neural-operator benchmarks, distribution-shift experiments, independent reference-solver validation, and representation-oracle diagnostics. The included files are sufficient to reproduce the reported numerical experiments, tables, and figures. All datasets are synthetically generated for this study; no human, clinical, observational, or third-party datasets are used.