Local privacy mechanisms preserve rate-double-robustness, enabling unbiased and semiparametrically efficient inference on target parameters indexed linearly by infinite-dimensional and nonlinearly by low-dimensional components from noisy private data.
Proceedings of the Third Conference on Theory of Cryptography , pages =
17 Pith papers cite this work, alongside 3,930 external citations. Polarity classification is still indexing.
17
Pith papers citing it
3,930
external citations · Crossref
citation-role summary
method 1
citation-polarity summary
roles
method 1polarities
use method 1representative citing papers
Gaussian mechanism is asymptotically optimal for high-dimensional DP additive noise; new Spherical Generalized Gamma family outperforms it and the ℓ2 mechanism in some low-dimensional cases with tight composition.
Fair fine-tuning under Equalized Odds yields a tight bound Adv(A, M_f) ≤ Δ_EO · W on adversarial advantage in distribution inference attacks, with empirical reductions below detection threshold across six datasets.
Postprocessing the discrete Laplace mechanism yields unbiased estimators for subexponential functions and equivalent distributions to Laplace or Staircase mechanisms under the same privacy parameters.
A technique for enforcing differential privacy in temporal runtime monitoring by analyzing dependencies and injecting noise into specifications while using tree mechanisms to limit accuracy loss.
Privatar uses horizontal frequency partitioning and distribution-aware minimal perturbation to enable private offloading of VR avatar reconstruction, supporting 2.37x more users with modest overhead.
Privacy and fairness cannot both be guaranteed in facility location over all datasets, but mechanisms exist that are optimal or near-optimal on welfare and fairness for natural data while preserving worst-case differential privacy.
Semantic Non-Assembly defines privacy via architectural inertness that blocks sub-threshold coalitions from assembling predicate inputs, with ProVerif verification of four properties in a two-channel architecture and a Birthmark Standard for constrained hardware.
DP4SQL enables customizable differentially private SQL for relational databases by supporting flexible policies for record existence, contents, partially public data, and varying protection levels across data parts.
CAPS provides an iterative differentially private synthesis method that outperforms one-shot baselines on authentic educational real-world data.
InvisibleInk achieves high-utility differentially private long-form LLM text generation at 4-8x the cost of non-private generation by isolating and clipping sensitive logits and sampling from a small superset of top-k private tokens without privacy cost.
The authors provide a detailed taxonomy of 21 risks associated with language models, covering discrimination, information leaks, misinformation, malicious applications, interaction harms, and societal impacts like job loss and environmental costs.
PPRE combines privacy technologies with Bayesian surname geocoding and survey data to enable production fairness measurements by race/ethnicity for U.S. LinkedIn members.
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
Proposes a formal DP-compatible framework with three unfairness measures (mutual information with TV proxy, MaxSAT-based repair, top-k tuple contribution) that satisfy positivity, monotonicity, and computability.
Empirical study of DP transfer learning reveals that larger clipping bounds outperform under tight privacy and cumulative DP noise explains batch-size effects better than existing heuristics.
citing papers explorer
No citing papers match the current filters.