Adversarial Robustness in Machine Learning

Últimos artículos

• Optimistic Policy Learning under Pessimistic Adversaries with Regret and Violation Guarantees

  Sourav Ganguly, Kartik Pandit, Arnob Ghosh · 17 de abril de 2026

  Real-world decision-making systems operate in environments where state transitions depend not only on the agent's actions, but also on \textbf{exogenous factors outside its control}--competing agents, environmental disturbances, or strategic adversaries--formally, $s_{h+1} = f(s_h, a_h, \bar{a}_h)+\…

• Retina gap junctions support the robust perception by warping neural representational geometries along the visual hierarchy

  Yang Yue, Shenjian Zhang, Yonghong Tian, Kai Du, Tiejun Huang · 17 de abril de 2026

  Deep Neural Networks (DNNs) are vulnerable to elaborately designed adversarial noise, although they have achieved extraordinary success in many tasks. Compared with DNNs, the human visual system is highly robust. However, it is unclear how the human visual system defends against adversarial attacks,…

• Hijacking Large Audio-Language Models via Context-Agnostic and Imperceptible Auditory Prompt Injection

  Meng Chen, Kun Wang, Li Lu, Jiaheng Zhang, Tianwei Zhang · 17 de abril de 2026

  Modern Large audio-language models (LALMs) power intelligent voice interactions by tightly integrating audio and text. This integration, however, expands the attack surface beyond text and introduces vulnerabilities in the continuous, high-dimensional audio channel. While prior work studied audio ja…

• Physically-Induced Atmospheric Adversarial Perturbations: Enhancing Transferability and Robustness in Remote Sensing Image Classification

  Weiwei Zhuang, Wangze Xie, Qi Zhang, Xia Du, Zihan Lin, Zheng Lin, Hanlin Cai, Jizhe Zhou, Zihan Fang, Chi-man Pun, Wei Ni, Jun Luo · 17 de abril de 2026

  Adversarial attacks pose a severe threat to the reliability of deep learning models in remote sensing (RS) image classification. Most existing methods rely on direct pixel-wise perturbations, failing to exploit the inherent atmospheric characteristics of RS imagery or survive real-world image degrad…

• RL-STPA: Adapting System-Theoretic Hazard Analysis for Safety-Critical Reinforcement Learning

  Steven A. Senczyszyn, Timothy C. Havens, Nathaniel Rice, Jason E. Summers, Benjamin D. Werner, Benjamin J. Schumeg · 17 de abril de 2026

  As reinforcement learning (RL) deployments expand into safety-critical domains, existing evaluation methods fail to systematically identify hazards arising from the black-box nature of neural network enabled policies and distributional shift between training and deployment. This paper introduces Rei…

• Calibrate-Then-Delegate: Safety Monitoring with Risk and Budget Guarantees via Model Cascades

  Edoardo Pona, Milad Kazemi, Mehran Hosseini, Yali Du, David Watson, Osvaldo Simeone, Nicola Paoletti · 17 de abril de 2026

  Monitoring LLM safety at scale requires balancing cost and accuracy: a cheap latent-space probe can screen every input, but hard cases should be escalated to a more expensive expert. Existing cascades delegate based on probe uncertainty, but uncertainty is a poor proxy for delegation benefit, as it …

• Context Over Content: Exposing Evaluation Faking in Automated Judges

  Manan Gupta, Inderjeet Nair, Lu Wang, Dhruv Kumar · 17 de abril de 2026

  The $\textit{LLM-as-a-judge}$ paradigm has become the operational backbone of automated AI evaluation pipelines, yet rests on an unverified assumption: that judges evaluate text strictly on its semantic content, impervious to surrounding contextual framing. We investigate $\textit{stakes signaling}$…

• Constrained Decoding for Safe Robot Navigation Foundation Models

  Parv Kapoor, Akila Ganlath, Michael Clifford, Changliu Liu, Sebastian Scherer, Eunsuk Kang · 17 de abril de 2026

  Recent advances in the development of robotic foundation models have led to promising end-to-end and general-purpose capabilities in robotic systems. Trained on vast datasets of simulated and real-world trajectories, these policies map multimodal observations directly to action sequences for physica…

• Beyond Uniform Sampling: Synergistic Active Learning and Input Denoising for Robust Neural Operators

  Samrendra Roy, Souvik Chakraborty, Syed Bahauddin Alam · 16 de abril de 2026

  Neural operators have emerged as fast surrogate models for physics simulations, yet they remain acutely vulnerable to adversarial perturbations, a critical liability for safety-critical digital twin deployments. We present a synergistic defense that combines active learning-based data generation wit…

• Activation-Guided Local Editing for Jailbreaking Attacks

  Jiecong Wang, Haoran Li, Hao Peng, Ziqian Zeng, Zihao Wang, Haohua Du, Zhengtao Yu · 16 de abril de 2026

  Jailbreaking is an essential adversarial technique for red-teaming these models to uncover and patch security flaws. However, existing jailbreak methods face significant drawbacks. Token-level jailbreak attacks often produce incoherent or unreadable inputs and exhibit poor transferability, while pro…

• Learning Probabilistic Responsibility Allocations for Multi-Agent Interactions

  Isaac Remy, Caleb Chang, Karen Leung · 16 de abril de 2026

  Human behavior in interactive settings is shaped not only by individual objectives but also by shared constraints with others, such as safety. Understanding how people allocate responsibility, i.e., how much one deviates from their desired policy to accommodate others, can inform the design of socia…

• WIN-U: Woodbury-Informed Newton-Unlearning as a retain-free Machine Unlearning Framework

  Xingjian Zhao, Mohammad Mohammadi Amiri, Malik Magdon-Ismail · 16 de abril de 2026

  Privacy concerns in LLMs have led to the rapidly growing need to enforce a data's "right to be forgotten". Machine unlearning addresses precisely this task, namely the removal of the influence of some specific data, i.e., the forget set, from a trained model. The gold standard for unlearning is to p…

• From Anchors to Supervision: Memory-Graph Guided Corpus-Free Unlearning for Large Language Models

  Wenxuan Li, Zhenfei Zhang, Mi Zhang, Geng Hong, Mi Wen, Xiaoyu You, Min Yang · 16 de abril de 2026

  Large language models (LLMs) may memorize sensitive or copyrighted content, raising significant privacy and legal concerns. While machine unlearning has emerged as a potential remedy, prevailing paradigms rely on user-provided forget sets, making unlearning requests difficult to audit and exposing s…

• Enhancing Confidence Estimation in Telco LLMs via Twin-Pass CoT-Ensembling

  Anton Saenko, Pranshav Gajjar, Abiodun Ganiyu, Vijay K. Shah · 16 de abril de 2026

  Large Language Models (LLMs) are increasingly applied to complex telecommunications tasks, including 3GPP specification analysis and O-RAN network troubleshooting. However, a critical limitation remains: LLM-generated confidence scores are often biased and unreliable, frequently exhibiting systemati…

• SafeHarness: Lifecycle-Integrated Security Architecture for LLM-based Agent Deployment

  Xixun Lin, Yang Liu, Yancheng Chen, Yongxuan Wu, Yucheng Ning, Yilong Liu, Nan Sun, Shun Zhang, Bin Chong, Chuan Zhou, Yanan Cao, Li Guo · 16 de abril de 2026

  The performance of large language model (LLM) agents depends critically on the execution harness, the system layer that orchestrates tool use, context management, and state persistence. Yet this same architectural centrality makes the harness a high-value attack surface: a single compromise at the h…

• Making AI Compliance Evidence Machine-Readable

  Rodrigo Cilla Ugarte, Miguel \'Angel Patricio Guisado, Antonio Berlanga de Jes\'us, Jos\'e Manuel Molina L\'opez · 16 de abril de 2026

  AI Assurance -- producing the machine-readable evidence required to demonstrate compliance with AI governance frameworks -- has mature policy scaffolding but lacks the infrastructure to operationalize it. Organizations building high-risk AI systems under the EU AI Act face a gap: frameworks such as …

• Who Gets Flagged? The Pluralistic Evaluation Gap in AI Content Watermarking

  Alexander Nemecek, Osama Zafar, Yuqiao Xu, Wenbiao Li, Erman Ayday · 16 de abril de 2026

  Watermarking is becoming the default mechanism for AI content authentication, with governance policies and frameworks referencing it as infrastructure for content provenance. Yet across text, image, and audio modalities, watermark signal strength, detectability, and robustness depend on statistical …

• The Cognitive Circuit Breaker: A Systems Engineering Framework for Intrinsic AI Reliability

  Jonathan Pan · 16 de abril de 2026

  As Large Language Models (LLMs) are increasingly deployed in mission-critical software systems, detecting hallucinations and ``faked truthfulness'' has become a paramount engineering challenge. Current reliability architectures rely heavily on post-generation, black-box mechanisms, such as Retrieval…

• HINTBench: Horizon-agent Intrinsic Non-attack Trajectory Benchmark

  Jiacheng Wang, Jinchang Hou, Fabian Wang, Ping Jian, Chenfu Bao, Zhonghou Lv · 16 de abril de 2026

  Existing agent-safety evaluation has focused mainly on externally induced risks. Yet agents may still enter unsafe trajectories under benign conditions. We study this complementary but underexplored setting through the lens of \emph{intrinsic} risk, where intrinsic failures remain latent, propagate …

• Monte Carlo Stochastic Depth for Uncertainty Estimation in Deep Learning

  Adam T. M\"uller, Tobias R\"ogelein, Nicolaj C. Stache · 15 de abril de 2026

  The deployment of deep neural networks in safety-critical systems necessitates reliable and efficient uncertainty quantification (UQ). A practical and widespread strategy for UQ is repurposing stochastic regularizers as scalable approximate Bayesian inference methods, such as Monte Carlo Dropout (MC…

• INTARG: Informed Real-Time Adversarial Attack Generation for Time-Series Regression

  Gamze Kirman Tokgoz, Onat Gungor, Tajana Rosing, Baris Aksanli · 15 de abril de 2026

  Time-series forecasting aims to predict future values by modeling temporal dependencies in historical observations. It is a critical component of many real-world systems, where accurate forecasts improve operational efficiency and help mitigate uncertainty and risk. More recently, machine learning (…

• LASA: Language-Agnostic Semantic Alignment at the Semantic Bottleneck for LLM Safety

  Junxiao Yang, Haoran Liu, Jinzhe Tu, Jiale Cheng, Zhexin Zhang, Shiyao Cui, Jiaqi Weng, Jialing Tao, Hui Xue, Hongning Wang, Han Qiu, Minlie Huang · 15 de abril de 2026

  Large language models (LLMs) often demonstrate strong safety performance in high-resource languages, yet exhibit severe vulnerabilities when queried in low-resource languages. We attribute this gap to a mismatch between language-agnostic semantic understanding ability and language-dominant safety al…

• GF-Score: Certified Class-Conditional Robustness Evaluation with Fairness Guarantees

  Arya Shah, Kaveri Visavadiya, Manisha Padala · 15 de abril de 2026

  Adversarial robustness is essential for deploying neural networks in safety-critical applications, yet standard evaluation methods either require expensive adversarial attacks or report only a single aggregate score that obscures how robustness is distributed across classes. We introduce the \emph{G…

• Efficient Adversarial Training via Criticality-Aware Fine-Tuning

  Wenyun Li, Zheng Zhang, Dongmei Jiang, Yaowei Wang, Xiangyuan Lan · 15 de abril de 2026

  Vision Transformer (ViT) models have achieved remarkable performance across various vision tasks, with scalability being a key advantage when applied to large datasets. This scalability enables ViT models to exhibit strong generalization capabilities. However, as the number of parameters increases, …

• TEMPLATEFUZZ: Fine-Grained Chat Template Fuzzing for Jailbreaking and Red Teaming LLMs

  Qingchao Shen, Zibo Xiao, Lili Huang, Enwei Hu, Yongqiang Tian, Junjie Chen · 15 de abril de 2026

  Large Language Models (LLMs) are increasingly deployed across diverse domains, yet their vulnerability to jailbreak attacks, where adversarial inputs bypass safety mechanisms to elicit harmful outputs, poses significant security risks. While prior work has primarily focused on prompt injection attac…