Search NASA⌕ Search

SEARCH · Search NASA

Results for “Adversarial examples”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Persistent Classification: Understanding Adversarial Attacks by Studying Decision Boundary Dynamics

ABSTRACT There are a number of hypotheses underlying the existence of adversarial examples for classification problems. These include the high‐dimensionality of the data, the high codimension in the ambient space of the data manifolds of interest, and that the structure of machine learning models may encourage classifiers to develop decision boundaries close to data points. This article proposes a new framework for studying adversarial examples that does not depend directly on the distance to the decision boundary. Similarly to the smoothed classifier literature, we define a (natural or adversarial) data point to be ( γ , σ)‐stable if the probability of the same classification is at least for points sampled in a Gaussian neighborhood of the point with a given standard deviation . We focus on studying the differences between persistence metrics along interpolants of natural and adversarial points. We show that adversarial examples have significantly lower persistence than natural examples for large neural networks in the context of the MNIST and ImageNet datasets. We connect this lack of persistence with decision boundary geometry by measuring angles of interpolants with respect to decision boundaries. Finally, we connect this approach with robustness by developing a manifold alignment gradient metric and demonstrating the increase in robustness that can be achieved when training with the addition of this metric.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Exploring Robust Features for Improving Adversarial Robustness

While deep neural networks (DNNs) have revolutionized many fields, their fragility to carefully designed adversarial attacks impedes the usage of DNNs in safety-critical applications. In this article, we strive to explore the robust features that are not affected by the adversarial perturbations, that is, invariant to the clean image and its adversarial examples (AEs), to improve the model’s adversarial robustness. Specifically, we propose a feature disentanglement model to segregate the robust features from nonrobust features and domain-specific features. Here, the extensive experiments on five widely used datasets with different attacks demonstrate that robust features obtained from our model improve the model’s adversarial robustness compared to the state-of-the-art approaches. Moreover, the trained domain discriminator is able to identify the domain-specific features from the clean images and AEs almost perfectly. This enables AE detection without incurring additional computational costs. With that, we can also specify different classifiers for clean images and AEs, thereby avoiding any drop in clean image accuracy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Semantic Stealth: Crafting Covert Adversarial Patches for Sentiment Classifiers Using Large Language Models

Deep learning models have been shown to be vulnerable to adversarial attacks, in which perturbations to their inputs cause the model to produce incorrect predictions. As opposed to adversarial attacks in computer vision, where small changes introduced to pixel values can drastically alter a model's output while remaining imperceptible to humans, text-based attacks are difficult to conceal due to the discrete nature of tokens. Consequently, unconstrained gradient-based attacks often produce adversarial examples that lack semantic meaning, rendering them detectable through visual inspection or perplexity filters. In contrast to methods that rely on gradient-based optimization in the embedding space, we propose an approach that leverages a Large Language Model's ability to generate grammatically correct and semantically meaningful text to craft adversarial patches that seamlessly blend in with the original input text. These patches can be used to alter the behavior of a target model, such as a text classifier. Since our approach does not rely on gradient backpropagation, it only requires access to the target model's confidence scores, making it a grey-box attack. We demonstrate the feasibility of our approach using open-source LLMs, including Intel's Neural Chat, Llama2, and Mistral-Instruct, to generate adversarial patches capable of altering the predictions of a distilBERT model fine-tuned on the IMDB reviews dataset for sentiment classification.

Roa Carvajal, Maria↗
DOE OSTI2025

Topological Signatures of Adversaries in Multimodal Alignments

Topological Data Analysis for Adversarial Detection (LANL O4937) - Detects adversarial examples in vision-language models using persistent homology and two-sample testing. Combines TDA features from CLIP embeddings with statistical methods (ME, SCF, SAMMD, C2ST) for robust detection across ImageNet, CIFAR-10/100.

Bhattarai, Manish↗
DOE OSTI2023

Evasion Attacks

Evasion Attacks are cases when an adversary is trying to hide something, or fool, a machine learning system. See reference here: Goodfellow, Shlens, Szegedy, Explaining and Harnessing Adversarial Examples. You can think of evasion attack as similar the little boy from the story, “A Wolf in Sheep’s Clothing”. The little boy appears as one thing (a wolf), but is actually just a little boy. Below are a few more examples that have appeared in literature and the media.

97 MATHEMATICS AND COMPUTING↗
NASA NTRSDate not supplied

NEUROSPF: A Tool For the Symbolic Analysis of Neural Networks

This paper presents NEUROSPF, a tool for the symbolic analysis of neural networks. Given a trained neural network model, the tool extracts the architecture and model parameters and translates them into a Java representation that is amenable for analysis using the Symbolic PathFinder symbolic execution tool. Notably, NEUROSPF encodes specialized peer classes for parsing the model’s parameters, thereby enabling efficient analysis. With NEUROSPF the user has the flexibility to specify either the inputs or the network internal parameters as symbolic, promoting the application of program analysis and testing approaches from software engineering to the field of machine learning. For instance, NEUROSPF can be used for coverage-based testing and test generation, finding adversarial examples and also constraint-based repair of neural networks, thus improving the reliability of neural networks and of the applications that use them.

neural networks↗
DOE OSTI2025

Model-based Hierarchical Reinforcement Learning for Improved Physical Security Design: A Prototype

Prior work in FY24 developed an adversarial AI agent aid in path analysis of physical protection systems. This agent, trained using a model-based reinforcement learning algorithm, was able to successfully learn the most vulnerable path in facilities. It was able to extend the current state of practice for physical protection design by exhibiting dynamic behavior based on current environmental conditions. Whereas PathTrace largely performs a static, graph-based analysis, the AI agent was able to make decisions based on relative position in the facility, current conditions (was the adversarial agnet discovered?), and proximity to secondary targets. The agent demonstrated some novel capabilities, but had limitations that need to be resolved before it can be used for production purposes. For example, the adversarial agent generalizes poorly and takes a relatively long time to train. Nonetheless, there is still considerable promise for developing the adversarial agent further in order to explore even richer, more dynamic behaviors (e.g., adversary motivations, environmental debris, and more). This work considers a complementary idea; development of a planning agent. The planning agent is envisioned as an auto-complete-like tool that can help accelerate security system design by human experts. The agent would respect existing barriers and sensors placed by a human expert while offering cost-effective suggestions (i.e., implicitly balancing effectiveness with cost) to improve the design. The goal is for this agent to be part of an expert’s toolbox, not to totally upend the current state-of-practice, or to displace human experts. The ultimate goal would be concurrent training of both the adversarial and planning agent together, to learn entirely through self-play. This would represent an entirely new way of performing system deign. We selected a hierarchical, model-based reinforcement learning algorithm to serve as the planning agent. This is an extension of concepts used in the prior FY24 adversarial agent work. There, we had a single agent acting an environment. Here, we have two different sub-agents (policies), working together, to form a complete agent. There is a manager policy, which can select abstract goals on slower time scales, and a worker, which performs primitive actions to reach goals selected by the manager. It is worth noting that this class of algorithm is challenging to work with. From our understanding, our work is one of the first successful uses of model-based reinforcement learning (MBRL) in nuclear energy1 , and likely the first hierarchical model-based reinforcement learning application in nuclear energy. Further, this work is one of the first known attempts to apply AI to perform a design tasks in nuclear energy. Consequently, there were significant implementation challenges and the bulk of the work was focused on successful implementation and algorithm design. The results presented here are very low technology readiness level as a consequence of the lack of related literature, but still represent a significant step forward in the pursuit of applied AI for design.

42 ENGINEERING↗
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 OSTI2024

Radio Frequency Spectrum Audit to Inventory Private Cellular Base Station Infrastructure

The ever-changing cellular communication landscape makes it difficult to identify, map, and localize cellular base stations. Localizing cellular base stations provides various advantages, including information security, cybersecurity, spectrum management, and interference detection. For example, the MITRE ATT&CK® (Adversarial Tactics, Techniques, and Common Knowledge architecture) [1] and Common Attack Pattern Enumeration and Classification [2] emphasize the importance of being able to minimize the cyber security threat presented by unregulated private cellular base stations (PCBS). The majority of published research looks at the malicious use of PCBSs and focuses on using data retrieved from user equipment (UE), data obtained from an application on the UE, or data shared between the UE and a mobile network to locate it. This innovative strategy, however, focuses on the passively discovered uniqueness of radio frequency (RF) transmissions from commercial cellular infrastructure received in a designated monitoring position (DMP).

42 ENGINEERING↗
DOE OSTI2024

Generative $β$-hairpin design using a residue-based physicochemical property landscape

De novo peptide design is a new frontier that has broad application potential in the biological and biomedical fields. Most existing models for de novo peptide design are largely based on sequence homology that can be restricted based on evolutionarily derived protein sequences and lack the physicochemical context essential in protein folding. Generative machine learning for de novo peptide design is a promising way to synthesize theoretical data that are based on, but unique from, the observable universe. In this study, we created and tested a custom peptide generative adversarial network intended to design peptide sequences that can fold into the -hairpin secondary structure. This deep neural network model is designed to establish a preliminary foundation of the generative approach based on physicochemical and conformational properties of 20 canonical amino acids, for example, hydrophobicity and residue volume, using extant structure-specific sequence data from the PDB. The beta generative adversarial network model robustly distinguishes secondary structures of hairpin from α helix and intrinsically disordered peptides with an accuracy of up to 96% and generates artificial -hairpin peptide sequences with minimum sequence identities around 31% and 50% when compared against the current NCBI PDB and nonredundant databases, respectively. These results highlight the potential of generative models specifically anchored by physicochemical and conformational property features of amino acids to expand the sequence-to-structure landscape of proteins beyond evolutionary limits.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Organizational Influence on Supply Chain for Digital Energy Infrastructure: Business Models, and Policy Landscape

Significant investment is driving the essential modernization and digitization of U.S. energy infrastructure, but the United States faces a key challenge in this grid transformation: our renewable and clean energy supply chains have limited capacity to source necessary digital assets through U.S. or allied sources. Batteries and their associated power electronic interfaces are key components to delivering clean and more resilient energy delivery, providing much-needed fast ramping, emergency discharge, generation, and operations support to the electric grid. These services have grown to be invaluable over the past 10 years and will soon be an irreplaceable component of energy delivery. While there have been significant strides to move supply chains for raw and critical materials to U.S. and allied nations, the control and power electronic industry has lagged, in part because of lower cost margins. For example, the United States now has growing capacity to manufacture solar photovoltaic (PV) panels, but 90% of the inverters—which are essential to the conversion of DC to AC for grid connection and controls—are made in or source parts from the People’s Republic of China (PRC). While a global supply chain is beneficial for many economic and supply diversification reasons, presence of foreign entities of concern (FEOC) in a dominant role in this supply chain brings additional concerns for national security of infrastructure. The BESS relationship and new clean energy have features that warrant discussions and additional focus on impact for technology security, given their critical role. This review considers the impact and solutions for complex business and policy landscapes in clean energy supply chain, and the benefit trade-offs, focusing on the Battery Energy Storage System case study, and its complex organizational relationships across both the US, Allied countries and Foreign Entities of Concern. While many examples in this paper are presented on PRC focused manufacture, this could apply to any potentially adversarial relationship

25 - ENERGY STORAGE↗
NASA NTRS2013

AGATE: Adversarial Game Analysis for Tactical Evaluation

AGATE generates a set of ranked strategies that enables an autonomous vehicle to track/trail another vehicle that is trying to break the contact using evasive tactics. The software is efficient (can be run on a laptop), scales well with environmental complexity, and is suitable for use onboard an autonomous vehicle. The software will run in near-real-time (2 Hz) on most commercial laptops. Existing software is usually run offline in a planning mode, and is not used to control an unmanned vehicle actively. JPL has developed a system for AGATE that uses adversarial game theory (AGT) methods (in particular, leader-follower and pursuit-evasion) to enable an autonomous vehicle (AV) to maintain tracking/ trailing operations on a target that is employing evasive tactics. The AV trailing, tracking, and reacquisition operations are characterized by imperfect information, and are an example of a non-zero sum game (a positive payoff for the AV is not necessarily an equal loss for the target being tracked and, potentially, additional adversarial boats). Previously, JPL successfully applied the Nash equilibrium method for onboard control of an autonomous ground vehicle (AGV) travelling over hazardous terrain.

Huntsberger, Terrance L.↗
DOE OSTI2024

Moment extraction using an unfolding protocol without binning

Deconvolving (“unfolding”) detector distortions is a critical step in the comparison of cross-section measurements with theoretical predictions in particle and nuclear physics. However, most existing approaches require histogram binning while many theoretical predictions are at the level of statistical moments. We develop a new approach to directly unfold distribution moments as a function of another observable without having to first discretize the data. Our moment unfolding technique uses machine learning and is inspired by Boltzmann weight factors and generative adversarial networks (GANs). We demonstrate the performance of this approach using jet substructure measurements in collider physics. With this illustrative example, we find that our moment unfolding protocol is more precise than bin-based approaches and is as or more precise than completely unbinned methods.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Application of Quantum Machine Learning to High Energy Physics Analysis at LHC using IBM Quantum Computer Simulators and IBM Quantum Computer Hardware

Our group pioneers the use of Quantum Machine Learning (QML) on High Energy Physics analysis at LHC. We have successfully employed several QML classification algorithms in the ttH (Higgs production in association with a top quark pair) and Higgs to two muons (Higgs coupling to second generation fermions), two recent LHC flagship physics analysis, on gate-model quantum computer simulators and hardware. The simulation studies have been performed with the IBM Quantum Framework, Google Tensorflow Quantum Framework, and Amazon Braket Framework, and we have achieved good classification performance that is similar to the performances of the classical machine learning methods currently used in LHC physics analyses, classical SVM, classical BDT, and classical deep neural network for example. We have also performed our studies using IBM superconducting quantum computer hardware and the performance is promising and is approaching the performance from IBM quantum simulators. Moreover, we extend our studies to other QML areas such as quantum anomaly detection and quantum generative adversarial, and some preliminary results have been obtained. Also, we have overcome the challenges of intensive computing resources in the cases of large qubits (25 qubits or more) and large numbers of events using NVIDIA cuQuantum with NERSC Perlmutter HPC. Our studies give an example that Quantum Machine Learning performs as well as its classical counterpart for realistic High Energy Physics analysis datasets. Furthermore, our result on noisy quantum hardware provides important validation for the result on noiseless quantum simulators.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Modeling Cyber Supply Chain Incidents with Multilayered Graph Motifs

As noted within the literature, supply chain includes people and organizations---manufacturers, integrators, and third-party vendors---that are involved in one or more stages of a product lifecycle. Since supply chains, by definition, include organizations and people, supply chain risk management activities must consider dependencies between an organization's business processes and third-party resources. Just as adversarial tactics can be implemented via techniques implemented via networked computer systems, so can such tactics be expressed via legal business relationships. A cyber incident may have an exponential impact downstream, for example, by leveraging a product's distribution channel (e.g. malicious updates in SolarWinds, buggy updates in CrowdStrike). Similarly, legitimate and legal business relationships also affect the attack surface exposure of systems, enabling long-term persistence and/or unknown impacts to product quality that are hard to detect. This paper catalogs several recent digital supply chain incidents and applies a multilayered network formalism to develop structural indicators (graph motifs) that reflect potentially-adversarial behavior. Finally, we compare and contrast the characteristics of adversarial tactics (e.g. Loss of Availability, Data Collection) that leverage cyber-physical dependencies to those that leverage legal organizational relationships.

97 - MATHEMATICS AND COMPUTING↗
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 OSTI2024

Towards Provable Security in Industrial Control Systems Via Dynamic Protocol Attestation

Industrial control systems (ICSs) increasingly rely on digital technologies vulnerable to cyber attacks. Cyber attackers can infiltrate ICSs and execute malicious actions. Individually, each action seems innocuous. But taken together, they cause the system to enter an unsafe state. These attacks have resulted in dramatic consequences such as physical damage, economic loss, and environmental catastrophes. This paper introduces a methodology that restricts actions using protocols. These protocols only allow safe actions to execute. Protocols are written in a domain specific language we have embedded in an interactive theorem prover (ITP). The ITP enables formal, machine-checked proofs to ensure protocols maintain safety properties. We use dynamic attestation to ensure ICSs conform to their protocol even if an adversary compromises a component. Since protocol conformance prevents unsafe actions, the previously mentioned cyber attacks become impossible. We demonstrate the effectiveness of our methodology using an example from the Fischertechnik Industry 4.0 platform. We measure dynamic attestation's impact on latency and throughput. Our approach is a starting point for studying how to combine formal methods and protocol design to thwart attacks intended to cripple ICSs.

97 MATHEMATICS AND COMPUTING↗