No article text was recorded for this version.
theoretical-empirical privacy gap
Version 2 of 2405 papers566 conceptsfrozen 2026-10-04
Papers in this version
- Blog: That which we call private
- Blog: The Elusive Pursuit of Reproducing PATE-GAN: Benchmarking, Auditing, Debugging
- Blog: Revealing the True Cost of Local Privacy: An Auditing Perspective
- Blog: Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning
- Blog: Auditing Differentially Private Machine Learning: How Private is Private SGD?
- Blog: A General Framework for Auditing Differentially Private Machine Learning
- Blog: Detecting Violations of Differential Privacy
- Blog: Auditing Differentially Private Machine Learning: How Private is Private SGD?
- Blog: Auditing $f$-differential privacy in one run
- Blog: Analyzing Privacy Leakage in Machine Learning via Multiple Hypothesis Testing: A Lesson From Fano
- Blog: DP-Auditorium: A Large-Scale Library for Auditing Differential Privacy
- Blog: The Elusive Pursuit of Replicating PATE-GAN: Benchmarking, Auditing, Debugging
- Blog: Bayesian Estimation of Differential Privacy
- Blog: It's Our Loss: No Privacy Amplification for Hidden State DP-SGD With Non-Convex Loss
- Blog: The Last Iterate Advantage: Empirical Auditing and Principled Heuristic Analysis of Differentially Private SGD
- Blog: Evaluating Differentially Private Machine Learning in Practice
- Blog: When Relaxations Go Bad: "Differentially-Private" Machine Learning
- Blog: The Last Iterate Advantage: Empirical Auditing and Principled Heuristic Analysis of Differentially Private SGD
- Blog: Why Does Differential Privacy with Large Epsilon Defend Against Practical Membership Inference Attacks?
- Blog: Debugging Differential Privacy: A Case Study for Privacy Auditing
- Blog: Antipodes of Label Differential Privacy: PATE and ALIBI
- Blog: One-shot Empirical Privacy Estimation for Federated Learning
- Blog: Wide Network Learning with Differential Privacy
- Blog: Accuracy First: Selecting a Differential Privacy Level for Accuracy-Constrained ERM
- Blog: How Well Can Differential Privacy Be Audited in One Run?
- Blog: Revealing the True Cost of Locally Differentially Private Protocols: An Auditing Perspective
- Blog: Thinking Outside of the Differential Privacy Box: A Case Study in Text Privatization with Language Model Prompting
- Blog: Curator Attack: When Blackbox Differential Privacy Auditing Loses Its Power
- Blog: Bayesian Estimation of Differential Privacy
- Blog: Bayesian Estimation of Differential Privacy
- Blog: "What do you want from theory alone?" Experimenting with Tight Auditing of Differentially Private Synthetic Data Generation
- Blog: An Efficient DP-SGD Mechanism for Large Scale NLU Models
- Blog: CANIFE: Crafting Canaries for Empirical Privacy Measurement in Federated Learning
- Blog: Empirical Privacy Variance
- Blog: Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning
- Blog: Analysis, Design, and Implementation of a User-Friendly Differential Privacy Application
- Blog: Tighter Privacy Auditing of DP-SGD in the Hidden State Threat Model
- Blog: P-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models
- Blog: On the Robustness of Information-Theoretic Privacy Measures and Mechanisms
- Blog: PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models
- Blog: MACE: A Flexible Framework for Membership Privacy Estimation in Generative Models
- Blog: Accuracy First: Selecting a Differential Privacy Level for Accuracy Constrained ERM
- Blog: PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models
- Blog: Privacy Auditing with One (1) Training Run
- Blog: Fast and Memory Efficient Differentially Private-SGD via JL Projections
- Blog: Four accuracy bounds and one estimator for frequency estimation under local differential privacy
- Blog: Tight and Practical Privacy Auditing for Differentially Private In-Context Learning
- Blog: Revealing the True Cost of Locally Differentially Private Protocols: An Auditing Perspective
- Blog: Sequentially Auditing Differential Privacy
- Blog: Differentially Private Domain Discovery
- Blog: Privacy Auditing with One (1) Training Run
- Blog: From Differential Privacy to Bounds on Membership Inference: Less can be More
- Blog: Challenging Differential Privacy: The Case of Non-interactive Mechanisms
- Blog: Beyond the Worst Case: Extending Differential Privacy Guarantees to Realistic Adversaries
- Blog: Defending against Reconstruction Attacks with Rényi Differential Privacy
- Blog: Gradients Look Alike: Sensitivity is Often Overestimated in DP-SGD
- Blog: Auditing Private Prediction
- Blog: Data Poisoning against Differentially-Private Learners: Attacks and Defenses
- Blog: Reconciling privacy and accuracy in AI for medical imaging
- Blog: Differentially Private Generalized Linear Models Revisited
Showing 60 of 405 papers.
History & citation
Current · Permanent link
- Empirical Privacy Audits for Machine Learning merged into this direction (100% overlap)
- Auditing Machine Learning Privacy through Empirical Attacks merged into this direction (59% overlap)
- Auditing and Verifying Differential Privacy Implementations merged into this direction (61% overlap)
- Single-Run Privacy Auditing for Large-Scale Models merged into this direction (100% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_theoretical_empirical_privacy_gap_df8fb4cb,
title = {theoretical-empirical privacy gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/theoretical-empirical-privacy-gap/rdr_df8fb4cb55e84188a52d044e64848123},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 2, frozen text):
@misc{lacuna_theoretical_empirical_privacy_gap_df8fb4cb_v2,
title = {theoretical-empirical privacy gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/theoretical-empirical-privacy-gap/rdr_df8fb4cb55e84188a52d044e64848123/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-20},
}