This paper shows that under standard assumptions, there is no polynomial-time verifier for exact verification of ReLU networks in an adversarial smoothed model.

The paper introduces SORA, an adaptive adversarial training method that dynamically adjusts perturbation sizes to prevent Catastrophic Overfitting, achieving state-of-the-art robustness and clean accu…

The paper introduces Confidence-Adaptive SwiGLU ($κ$-SwiGLU), a novel gating mechanism for Mixture-of-Experts (MoE) models that dynamically adjusts the gate sharpness based on token-level routing conf…

The paper proposes Sensitivity-Uncertainty Alignment (SUA), a framework that measures the misalignment between a model's prediction instability and its stated uncertainty to improve model reliability.

This paper provides the first non-vacuous generalization analysis for the Stochastic Variance Reduced Gradient (SVRG) method by establishing sharp, data-dependent algorithmic stability bounds, thereby…

The paper proposes a debiasing fine-tuning technique to efficiently enhance the robustness of Large Language Models against semantically similar but textually altered prompts.

The paper proposes a unified, architecture-agnostic framework that significantly improves the robustness of deepfake image detectors against adversarial attacks by focusing on higher-order frequency s…

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

This paper proposes using random sampling of prediction precision during inference to significantly enhance the adversarial robustness of Automatic Speech Recognition (ASR) systems.

This paper provides explicit error bounds for the infinite-width Gaussian-process limit of random neural networks using tensor programs and quantitative convergence theory in Wasserstein distance.

The paper proposes combining Gaussian noise and bilateral filtering into a simple preprocessor that achieves supralinear and scalable adversarial robustness in CNNs with significantly reduced computat…

The paper proposes a general-purpose pipeline to train automated red teaming models capable of generating attacks for arbitrary adversarial goals, overcoming the limitations of current methods that ar…

The paper demonstrates that adversarial examples can be used to manipulate Vision-Language Models (VLMs) into confidently providing authoritative but incorrect information, a process termed 'AI author…

The paper introduces Singularity-aware Adam (S-Adam), a novel optimizer that stabilizes deep learning training in non-smooth loss landscapes by dynamically damping updates based on local geometric ins…

The paper proposes a novel framework using the primal-dual perspective of differential privacy to provide a unified, modular, and end-to-end robustness certification for complex machine learning model…

This paper studies the behavior of mixture-of-experts (MoE) models under distribution shift and proposes an adversarial reweighting method to improve their calibration.

This paper tests the impact of acoustic factors on voice control systems and introduces a Dual-Form Signal to Noise Ratio to decouple source stealth from attack efficacy.

The paper proposes a Distribution-aware Dynamic Guidance (DDG) strategy to mitigate catastrophic overfitting and the robustness-accuracy trade-off inherent in Fast Adversarial Training (FAT) by dynami…

This study longitudinally evaluates the adversarial robustness of Android malware detection systems over a decade, finding that temporal separation significantly degrades robustness due to concept dri…

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.