From Pixels to Privacy-Sensitive Features: An Information-Theoretic Framework for Reconstruction Bounds

The paper presents an information‑theoretic framework that derives a distortion bound for reconstruction attacks on models trained with DP‑SGD. The bound can be computed before training and applies even when an adversary observes the training trajectory and knows all other training data. It separates the information leaked by training (conditional mutual information) from the inherent uncertainty in the data (source entropy) and uses rate‑distortion theory to produce a lower bound on expected reconstruction error, instantiated in both pixel and LPIPS feature spaces.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

• safety