BOLT: Boolean-feature One-release Label Transfer for Differentially Private Tabular Synthetic Data

Differentially private tabular synthesis must balance predictive utility with the computational cost of producing usable data. We introduce BOLT (Boolean-feature One-release Label Transfer), which constructs two complementary models from a single jointly calibrated Gaussian release: a simple class-conditional covariate sampler and a richer Bernoulli na\"ive-Bayes classifier over fixed threshold predicates and randomly selected two-literal conjunctions. BOLT samples covariates from the simpler model and assigns stochastic labels using the classifier, incorporating private interaction statistics into prediction without fitting a generator to reproduce those statistics. Under a fixed public schema and encoding, synthesis is postprocessing of the shared $(\epsilon ,\delta )$-differentially private release. We evaluate five private synthesis methods across five datasets, four privacy budgets ($\epsilon \in \{0.5,1,2,8\}$), and two downstream classifiers. Averaging F1 and AUROC ratios to the corresponding nonprivate real-data references equally across datasets, budgets, learners, and metrics, BOLT retains 87.0% of reference utility with a mean complete-table synthesis time of $2.85$ seconds, compared with 89.3% and $5.16$ hours for AIM. BOLT achieves the lowest mean synthesis time in all twenty dataset–privacy configurations. Ratios of recorded arithmetic mean runtimes give $12.9\times$ and $6,517\times$ speed advantages over DP-MERF, the next-fastest method, and AIM, respectively. After averaging scores across releases, BOLT exceeds each of DP-MERF, DP-HP, and Tab-PE in at least 80% of the dataset–budget–learner–metric comparisons.