Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “adversarial attack”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

XploreNAS : Explore Adversarially Robust and Hardware-efficient Neural Architectures for Non-ideal Xbars

Compute In-Memory platforms such as memristive crossbars are gaining focus as they facilitate acceleration of Deep Neural Networks (DNNs) with high area and compute efficiencies. However, the intrinsic non-idealities associated with the analog nature of computing in crossbars limits the performance of the deployed DNNs. Furthermore, DNNs are shown to be vulnerable to adversarial attacks leading to severe security threats in their large-scale deployment. Thus, finding adversarially robust DNN architectures for non-ideal crossbars is critical to the safe and secure deployment of DNNs on the edge. This work proposes a two-phase algorithm-hardware co-optimization approach called XploreNAS that searches for hardware efficient and adversarially robust neural architectures for non-ideal crossbar platforms. We use the one-shot Neural Architecture Search approach to train a large Supernet with crossbar-awareness and sample adversarially robust Subnets therefrom, maintaining competitive hardware efficiency. Our experiments on crossbars with benchmark datasets (SVHN, CIFAR10, CIFAR100) show up to ~8–16% improvement in the adversarial robustness of the searched Subnets against a baseline ResNet-18 model subjected to crossbar-aware adversarial training. We benchmark our robust Subnets for Energy-Delay-Area-Products (EDAPs) using the Neurosim tool and find that with additional hardware efficiency–driven optimizations, the Subnets attain ~1.5–1.6× lower EDAPs than ResNet-18 baseline.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Dynamic Low-Rank Training with Spectral Regularization: Achieving Robustness in Compressed Representations

Deployment of neural networks on resource-constrained devices demands models that are both compact and robust to adversarial inputs. However, compression and adversarial robustness often conflict. In this work, we introduce a dynamical low-rank training scheme enhanced with a novel spectral regularizer that controls the condition number of the low-rank core in each layer. This approach mitigates the sensitivity of compressed models to adversarial perturbations without sacrificing clean accuracy. The method is model- and data-agnostic, computationally efficient, and supports rank adaptivity to automatically compress the network at hand. Extensive experiments across standard architectures, datasets, and adversarial attacks show the regularized networks can achieve over 94\% compression while recovering or improving adversarial accuracy relative to uncompressed baselines.

Schotthoefer, Steffen [ORNL] (ORCID:00000002156965↗
DOE OSTI2024

A Compound Data Poisoning Technique with Significant Adversarial Effects on Transformer-based Sentiment Classification Tasks

Transformer-based models have demonstrated much success in various natural language processing tasks. However, they are often vulnerable to adversarial attacks, such as data poisoning, which can intentionally fool the model into generating incorrect results. In this article, we present a novel, compound variant of a data poisoning attack on a transformer-based model that maximizes the poisoning effect while minimizing the scope of poisoning. Here we do so by combining the established data poisoning technique (label flipping) with a novel adversarial artifact selection and insertion technique aimed at minimizing detectability and the scope of the poisoning footprint. We find that by using a combination of these two techniques, we achieve a state-of-the-art attack success rate of approximately 90% while poisoning only 0.5% of the original training set, thus minimizing the scope and detectability of the poisoning action. These findings have the potential to advance the development of better data poisoning detection methods.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Emerging Threats in Transportation Security Related to Intelligent Transportation Systems (ITS)

Transport of high-consequence shipments requires a resilient and robust systems of systems to guarantee cargo arrival. Furthermore, rising adoption of technologies such as connected and automated vehicles (CAVs), intelligent infrastructure, and vehicle-to-everything (V2X) communication presents unique challenges for securing transportation systems. Within these Intelligent Transportation Systems (ITS), several additional vulnerabilities exist that create pathways for adversarial attacks and cargo interception. For example, connectivity provides cyber pathways directly into vehicle systems and infrastructure for malicious actors. Furthermore, advanced vehicle automation exposes additional vehicle control necessary for shipment interception otherwise unavailable to adversaries. Within this paper, we will discuss the specific threats introduced by ITS-enabled technologies currently deployed and in development. These include those mentioned related to connectivity and automation, but will be expanded into grid, infrastructure, and vehicle specific threats. In addition, we will discuss how to potentially mitigate these emerging challenges as well as how to safeguard transportation systems from next generation attacks.

Cook, Adian [ORNL] (ORCID:0000000160825395)↗
DOE OSTI2023

PathTrace and MPVEASI: A Path Analysis Comparative Validation Study

Developed in 2018, PathTrace is a software package built with the intention of making path analysis simple and intuitive. PathTrace is a top-down pathway analysis software where a user is able to explore vulnerable pathways into a facility. The intention of utilizing a software tool like PathTrace is to characterize an existing physical protection system (PPS) and to upgrade the system to achieve a high level of response interruption, or probability of interruption (P I ) of the adversary. There are four steps for conducting path analysis using PathTrace. The first step is to identify an image to use to build the model and scale the model within PathTrace using a section of known distance (wall or fence perimeter, for example). The scaling process will produce a grid of cells through which the user is able to build a model. The second step is to fill out the grid of cells with four categories of materials: Barriers, Detection Areas, Jumps, and Targets. These materials apply associated delay and detection values to the cells in which they are applied. The third step is to represent the adversary and response forces. The adversaries are represented by their capabilities in interacting with the materials identified in step two, and the response is represented by how quickly they will be able to respond to an adversary attack. Finally, the user is able to take all of the information from the previous three steps and perform a Most Vulnerable Path (MVP) analysis. In this stage, the user is able to visualize vulnerable adversary pathways and reason about how to upgrade these pathways to provide a high level of P I .

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

METHODOLOGY AND APPLICATION OF PHYSICAL SECURITY EFFECTIVENESS BASED ON DYNAMIC FORCE-ON-FORCE MODELING

This paper describes ongoing work within the Light Water Reactor Sustainability (LWRS) Program at Idaho National Laboratory (INL) to optimize security and cost of nuclear power plants (NPPs). It reviews the conservatisms in conventional physical security posture and regulations. It introduces the dynamic risk assessment tool developed at INL, Event Modeling Risk Assessment using Linked Diagrams (EMRALD). The dynamic assessment methodology leverages EMRALD to process results of force-on-force (FOF) simulations and crediting safety mitigation actions from probabilistic risk assessment (PRA) models as well as diverse and flexible coping strategies (FLEX) mitigation strategies. Timing information from these simulations are compared against the available time to perform mitigations obtained from Reactor Excursion and Leak Analysis Program (RELAP5) simulations. To illustrate the methodology, a station blackout (SBO) attack scenario was modeled in commercially available FOF simulation tools. The simulation results provide valuable insights into possible attack outcomes and as the probabilistic risk of a core damage event given these outcomes. Safety mitigation procedures were modeled in EMRALD, and were dependent on the attack outcomes by considering human operator uncertainties. RELAP5 simulations incorporating human and hardware uncertainties were performed to estimate the distribution of time-to-core damage. The results demonstrate that, even in the extreme case of a successful adversarial attack, plant mitigation strategies provide significantly high-likelihood of preventing radiological release. The proposed modeling and simulation framework of integrating FLEX equipment with FOF models enables the NPPs to credit FLEX portable equipment in the plant security posture, resulting in an efficient and optimized physical security.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2025

Robustness of Deep Learning Classification to Adversarial Input on GPUs: Asynchronous Parallel Accumulation Is a Source of Vulnerability

The ability of machine learning (ML) classification models to resist small, targeted input perturbations—known as adversarial attacks—is a key measure of their safety and reliability. We show that floating-point non associativity (FPNA) coupled with asynchronous parallel programming on GPUs is sufficient to result in misclassification, without any perturbation to the input. Additionally, we show that this misclassification is particularly significant for inputs close to the decision boundary and that standard adversarial robustness results may be overestimated up to 4.6 when not considering machine-level details. We first study a linear classifier, before focusing on standard Graph Neural Network (GNN) architectures and datasets used in robustness assessments. We develop a novel black-box attack using Bayesian optimization to discover external workloads that can change the instruction scheduling which bias the output of reductions on GPUs and reliably lead to misclassification. Motivated by these results, we present a new learnable permutation (LP) gradient-based approach to learning floating-point operation orderings that lead to misclassifications. The LP approach provides a worst-case estimate in a computationally efficient manner, avoiding the need to run identical experiments tens of thousands of times over a potentially large set of possible GPU states or architectures. Finally, using instrumentation-based testing, we investigate parallel reduction ordering across different GPU architectures under external background workloads, when utilizing multi-GPU virtualization, and when applying power capping. Our results demonstrate that parallel reduction ordering varies significantly across architectures under the first two conditions, substantially increasing the search space required to fully test the effects of this parallel scheduler-based vulnerability. These results and the methods developed here can help to include machine-level considerations into adversarial robustness assessments, which can make a difference in safety and mission critical applications.

Shanmugavelu, Sanjif [Maxeler Technologies, a Groq↗
DOE OSTI2022

Modeling design and control problems involving neural network surrogates

Here, we consider nonlinear optimization problems that involve surrogate models represented by neural networks. We demonstrate first how to directly embed neural network evaluation into optimization models, highlight a difficulty with this approach that can prevent convergence, and then characterize stationarity of such models. We then present two alternative formulations of these problems in the specific case of feedforward neural networks with ReLU activation: as a mixed-integer optimization problem and as a mathematical program with complementarity constraints. For the latter formulation we prove that stationarity at a point for this problem corresponds to stationarity of the embedded formulation. Each of these formulations may be solved with state-of-the-art optimization methods, and we show how to obtain good initial feasible solutions for these methods. We compare our formulations on three practical applications arising in the design and control of combustion engines, in the generation of adversarial attacks on classifier networks, and in the determination of optimal flows in an oil well network.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Quantum adversarial learning for kernel methods

We show that hybrid quantum classifiers based on quantum kernel methods and support vector machines are vulnerable against adversarial attacks, namely small engineered perturbations of the input data can deceive the classifier into predicting the wrong result. Nonetheless, we also show that simple defense strategies based on data augmentation with a few crafted perturbations can make the classifier robust against new attacks. Our results find applications in security-critical learning problems and in mitigating the effect of some forms of quantum noise, since the attacker can also be understood as part of the surrounding environment.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

Graph interpolating activation improves both natural and robust accuracies in data-efficient deep learning

Improving the accuracy and robustness of deep neural nets (DNNs) and adapting them to small training data are primary tasks in deep learning (DL) research. In this paper, we replace the output activation function of DNNs, typically the data-agnostic softmax function, with a graph Laplacian-based high-dimensional interpolating function which, in the continuum limit, converges to the solution of a Laplace–Beltrami equation on a high-dimensional manifold. Furthermore, we propose end-to-end training and testing algorithms for this new architecture. The proposed DNN with graph interpolating activation integrates the advantages of both deep learning and manifold learning. Compared to the conventional DNNs with the softmax function as output activation, the new framework demonstrates the following major advantages: First, it is better applicable to data-efficient learning in which we train high capacity DNNs without using a large number of training data. Second, it remarkably improves both natural accuracy on the clean images and robust accuracy on the adversarial images crafted by both white-box and black-box adversarial attacks. Third, it is a natural choice for semi-supervised learning. This paper is a significant extension of our earlier work published in NeurIPS, 2018. For reproducibility, the code is available at https://github.com/BaoWangMath/DNN-DataDependentActivation .

Mathematics↗
DOE OSTI2022

Prediction of Power Measurements Using Adaptive Filters

With the advent of smart grid concept, Internet of Things (IoT) and the deployment of smart meters, the cyberattack threats on power networks have increased due to the use of communication systems that can be accessed by adversaries. Attackers will have the ability to manipulate the outcomes of smart meters which in turn influence the core application of Energy Management System 9EMS): State Estimation (SE). Bad data analytic tools may fail to detect some attacks into measurements. Meanwhile, Machine Learning (ML) solutions have been proposed for detecting False Data Injection (FDI) attacks. However, there is a lack of ML time-series solutions presented in the state-of-the-art that is yet to be complex. In signal processing, time-series solutions do not only consider the signal, but also the statistics of the signal over time. Therefore, in this paper, a machine learning for time-series solutions is presented as an application to model the measurements of the power grid that are used in SE. The presented model takes into account adaptive linear and non-linear filters: Finite Impulse Response (FIR), and Infinite Impulse Response (IIR). The presented models are implemented and performed on the IEEE-118 bus system. The results indicate the advantage of applying those filters over the state-of-the-art machine learning solutions.

Hamad, Khaled↗
DOE OSTI2025

Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning

In this paper, we propose a learning-based method utilizing the Soft Actor-Critic (SAC) algorithm to train a binary Support Vector Machine (SVM) classifier. This classifier is designed to identify valid input spaces in high-dimensional, highly constrained systems while minimizing the total runtime of offline simulations. The simulations adapt their runtime based on the likelihood that a given training input will be informative to the classifier. Furthermore, we introduce a method for using the trained SAC model to predict whether a desired system input is likely to violate constraints, along with a technique to adjust the input as necessary. Additionally, we explore the potential of this model to detect faults or adversarial attacks within the system. The effectiveness of our approach is demonstrated through various simulations of challenging classification problems and a constrained quadrotor model.

Netter, Josh [Georgia Institute of Technology, Atl↗
DOE OSTI2021

Distributed State Estimation Over Time-Varying Graphs: Exploiting the Age-of-Information

Here, we study the problem of designing a distributed observer for an LTI system over a time-varying communication graph. The limited existing work on this topic imposes various restrictions either on the observation model or on the sequence of communication graphs. In contrast, we propose a single-time-scale distributed observer that works under mild assumptions. Specifically, our communication model only requires strong-connectivity to be preserved over non-overlapping, contiguous intervals that are even allowed to grow unbounded over time. We show that under suitable conditions that bound the growth of such intervals, joint observability is sufficient to track the state of any discrete-time LTI system exponentially fast, at any desired rate. We also develop a variant of our algorithm that is provably robust to worst-case adversarial attacks, provided the sequence of graphs is sufficiently connected over time. The key to our approach is the notion of a "freshness-index" that keeps track of the age-of-information being diffused across the network. Such indices enable nodes to reject stale estimates of the state, and, in turn, contribute to stability of the error dynamics.

42 ENGINEERING↗
DOE OSTI2024

Double Visual Defense

This is the official code for the paper "Double Visual Defense: Adversarial Pre-training and Instruction Tuning for Improving Vision-Language Model Robustness". This code can be used to produce vision language models (VLMs), like LLaVA, with enhanced robustness to adversarial attacks (e.g. jailbreaks).

Bartoldson, Brian [Lawrence Livermore National Lab↗
DOE OSTI2026

Evaluating lightweight unsupervised online IDS for masquerade attacks in CAN

Vehicular controller area networks (CANs) are susceptible to masquerade attacks by malicious adversaries. In masquerade attacks, adversaries silence a targeted ID and then send malicious frames with forged content at the expected timing of benign frames. As masquerade attacks could seriously harm vehicle functionality and are the stealthiest attacks to detect in CAN, recent work has devoted attention to compare frameworks for detecting masquerade attacks in CAN. However, most existing works report offline evaluations using CAN logs already collected using simulations that do not comply with the domain’s real-time constraints. Here we contribute to advance the state of the art by presenting a comparative evaluation of four different non-deep learning (DL)-based unsupervised online intrusion detection systems (IDS) for masquerade attacks in CAN. Our approach differs from existing comparative evaluations in that we analyze the effect of controlling streaming data conditions in a sliding window setting. In doing so, we use realistic masquerade attacks being replayed from the ROAD dataset. We show that although evaluated IDS are not effective at detecting every attack type, the method that relies on detecting changes in the hierarchical structure of clusters of time series produces the best results at the expense of higher computational overhead. We discuss limitations, open challenges, and how the evaluated methods can be used for practical unsupervised online CAN IDS for masquerade attacks.

Anomaly detection↗
DOE OSTI2021

Transfer Learning using Denoising Auto-Encoders for Cellular-Level Annotation of Tumor in Pathology Slides

Adversarial examples can produce altered classifications using only seemingly innocuous, imperceptible perturbations to the original image. The imperceptibility of adversarial perturbations suggests that the corresponding classifiers use decision criteria different than those of a human. In a medical setting, inexplicable decision criteria confound a pathologist’s willingness to trust machine-generated annotations. Here, we analyze denoising tumor detection models to see if they are robust to imperceptible adversarial perturbations. Moreover, to be more fully trusted by pathologists, we require tumor detectors that generate interpretable annotations which segment pathology slides into tumorous and normal regions at the cellular level. We therefore compare transfer learning based on two different autoencoder architectures, one derived from a deep denoising bottleneck autoencoder and one from an over-complete sparse autoencoder. Both autoencoders were first trained in an unsupervised manner on a set of pathology slides drawn from the Camelyon16 dataset. The latent representations produced by each autoencoder were then passed to separate neural networks that were trained in a supervised manner on binary tumor-normal masks generated by pathologists at cellular resolution. Both tumor detectors supported better than 90% AUC PR as measured by the area under the precision/recall curve on a held-out pathology slide. To assess the underlying decision criteria used by both tumor detectors, we constructed imperceptible adversarial examples which reduced the AUC PR of both models to less than 70%. Random noise of the same amplitude had almost no effect on the AUC PR of either model. Additionally, each tumor detector was resistant to adversarial “transfer” attacks targeting the other. The adversarial perturbations showed strong characteristic differences: the deep denoising models perturbations were a very diffuse, seemingly unrecognizable pattern while the sparse coding models perturbations showed traces of tissue cells.

47 OTHER INSTRUMENTATION↗