On Reconstructing Training Data From Bayesian Posteriors and Trained Models

Open in new window