Results 81 to 90 of about 15,235,426 (349)
Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching [PDF]
Data Poisoning attacks modify training data to maliciously control a model trained on such data. In this work, we focus on targeted poisoning attacks which cause a reclassification of an unmodified test image and as such breach model integrity. We consider a particularly malicious poisoning attack that is both "from scratch" and "clean label", meaning ...
arxiv
Explore the Effect of Data Selection on Poison Efficiency in Backdoor Attacks [PDF]
As the number of parameters in Deep Neural Networks (DNNs) scales, the thirst for training data also increases. To save costs, it has become common for users and enterprises to delegate time-consuming data collection to third parties. Unfortunately, recent research has shown that this practice raises the risk of DNNs being exposed to backdoor attacks ...
arxiv
2D Multifunctional Spin‐Orbit Coupled Dirac Nodal Line Materials
A total of 473 nonmagnetic and antiferromagnetic 2D spin‐orbit coupled Dirac nodal line materials are screened, spanning 5 layer groups and 12 magnetic space groups. Furthermore, it integrates their topological properties with electride, multiferroic, and magnetic characteristics, revealing unique systems with expanded functionalities and promising ...
Weizhen Meng+7 more
wiley +1 more source
Poisoned classifiers are not only backdoored, they are fundamentally broken [PDF]
Under a commonly-studied backdoor poisoning attack against classification models, an attacker adds a small trigger to a subset of the training data, such that the presence of this trigger at test time causes the classifier to always predict some target class.
arxiv
Schematic illustration of the mechanism of targeted delivery of nanoparticles including 1) paracellular, 2) endolysosomal escape, 3) receptor mediated endocytosis, and 4) M cell mediated transport for non‐GI diseases by oral administration such as atherosclerosis, cancer, diabetes, and brain diseases. Abstract Oral drug delivery is a promising approach
Subarna Ray+2 more
wiley +1 more source
Data Poisoning Attacks on Regression Learning and Corresponding Defenses [PDF]
Adversarial data poisoning is an effective attack against machine learning and threatens model integrity by introducing poisoned data into the training dataset. So far, it has been studied mostly for classification, even though regression learning is used in many mission critical systems (such as dosage of medication, control of cyber-physical systems ...
arxiv
Parity effect in superconducting aluminum single electron transistors with spatial gap profile controlled by film thickness [PDF]
We propose a novel method for suppression of quasiparticle poisoning in Al Coulomb blockade devices. The method is based on creation of a proper energy gap profile along the device. In contrast to the previously used techniques, the energy gap is controlled by the film thickness.
arxiv +1 more source
Propofol as a controlled substance: poison or remedy [PDF]
Korea is the first, and perhaps only, country to classify propofol as a controlled substance. In February 2011, the Korean Ministry of Food and Drug Safety classified propofol as a psychotropic drug, making it illegal to prescribe or consume other than for stipulated treatments that may need sedation, such as a gastrointestinal endoscopy [1].
openaire +4 more sources
Covalent Organic Frameworks for Photocatalysis
This review provides an overview of recent advances in covalent organic frameworks (COFs) for photocatalysis, focusing on sustainable energy applications like water splitting, hydrogen peroxide generation, and CO2 and N2 reduction. It discusses design principles, structure‐function relationships, challenges in COF photocatalysis, and strategies to ...
Bikash Mishra+6 more
wiley +1 more source
Laurel but Hardy: unintended poisoning, a case report of oleander misidentification as bay laurel
Introduction: Nerium oleander is a toxic plant containing cardiac glycosides throughout all its parts, thereby posing severe health risks upon ingestion.
Giovanni Mantelli+8 more
doaj