The paper proposes a certifiably robust malware detection framework using randomized smoothing and feature ablation to guarantee detection accuracy against metamorphic evasion attacks.

The paper proposes Byz-Clip21-SGD2M, a novel algorithm that achieves high-probability convergence guarantees for Federated Learning by integrating robust aggregation, double momentum, and clipping, re…

The paper introduces Tail-risk Intrinsic Geometric Smoothing (TIGS), a plug-and-play, inference-time defense that suppresses backdoor attacks on LLMs by structurally smoothing the attention mechanism…

This paper corrects the theoretical analysis of DP-SGD by identifying that common implementations, which use batch averaging, result in weaker privacy guarantees than previously reported.

An adversarial model trainer can plant statistically undetectable backdoors in deep neural networks, providing access to adversarial examples.

This paper proposes a density-aware attack that constructs triggers by placing poisoned samples in low-density regions of the clean data distribution, achieving high attack success rates even after st…

The paper introduces Disrupt-and-Rectify Smoothing (DR-Smoothing), a novel two-stage defense mechanism that significantly improves LLM security against jailbreaking attacks by restoring disrupted inpu…

The paper proposes a novel exponential mechanism using quadratic approximations to fine-tune machine learning models on sensitive data while providing strong differential privacy guarantees.

The paper introduces 'mixture mechanisms,' a novel class of additive noise mechanisms that achieve approximate differential privacy by mixing multiple Gaussian distributions, resulting in lower noise…

The paper introduces 'mixture mechanisms,' a novel class of additive noise mechanisms that achieve differential privacy for real-valued queries, significantly reducing noise compared to the standard G…

The paper proposes a Quantitative Information Flow (QIF) framework to systematically and rigorously compare Local Differential Privacy (LDP) frequency estimation protocols, moving beyond simple $\vare…

The paper develops a general framework to exactly characterize the composition of mechanisms satisfying multiple differential privacy constraints, extending known results to arbitrary numbers of const…

The paper introduces a lightweight, sampling-based cryptographic protocol for verifiable AI inference that drastically reduces proving overhead from minutes to milliseconds by leveraging statistical p…

The paper introduces an optimal black-box auditing framework using Donsker-Varadhan estimators to estimate Rényi differential privacy (RDP) guarantees for machine learning algorithms.

The paper introduces Zero-Run privacy auditing, a post-hoc framework that allows for practical differential privacy evaluation of large, deployed models without requiring retraining or controlled data…

The paper introduces the PML envelope, a novel definition that provides a robust and operationally meaningful measure of information leakage about a secret, satisfying both post-processing robustness…

TADP-RME introduces a trust-adaptive differential privacy framework that enhances data system reliability by dynamically adjusting the privacy budget based on user trust and disrupting geometric struc…

This paper proposes two post-processing techniques, random selection and linear combination, to construct a model that satisfies any desired differential privacy level without retraining, given a set…

This paper analyzes the reliability of efficient membership inference attack (MIA) evaluation methods, demonstrating that standard aggregation techniques introduce biases that compromise accurate vuln…

The paper proposes CEAR, an ensemble-based method that combines empirical and certified defenses to achieve superior provable robustness against adversarial attacks in Deep Neural Networks.