Alex VaccariCracking Black-Box Vision Models: Differential Evolution in ActionThis story is the second in a series on evolution-based attacks on computer vision models. If you haven’t already, be sure to check out…13h ago
InTowards Data SciencebyParul PandeyCan Recommendations from LLMs Be Manipulated to Enhance a Product’s Visibility?Impact of Large Language Models on E-CommerceApr 292
Ondra RojčíkThe XZ Backdoor: Unveiling the Elaborated Social Engineering Tactics of a Persistent Threat ActorThe XZ Backdoor incident highlights a significant infiltration of a critical open-source project through sophisticated social engineering…2d ago2d ago
Yoshi SatoUnderstanding FGSM: A more intuitive Approach to Adversarial AttacksUnderstand why machine learning models are so sensitive to subtle input changes through the lens of FGSM and high-dimensional geometry.2d ago2d ago
InThe GeneratorbyMoshe Sipper, Ph.D.Real, Fake, Attack?When deepfake detectors are attacked.Mar 132Mar 132
Alex VaccariCracking Black-Box Vision Models: Differential Evolution in ActionThis story is the second in a series on evolution-based attacks on computer vision models. If you haven’t already, be sure to check out…13h ago
InTowards Data SciencebyParul PandeyCan Recommendations from LLMs Be Manipulated to Enhance a Product’s Visibility?Impact of Large Language Models on E-CommerceApr 292
Ondra RojčíkThe XZ Backdoor: Unveiling the Elaborated Social Engineering Tactics of a Persistent Threat ActorThe XZ Backdoor incident highlights a significant infiltration of a critical open-source project through sophisticated social engineering…2d ago
Yoshi SatoUnderstanding FGSM: A more intuitive Approach to Adversarial AttacksUnderstand why machine learning models are so sensitive to subtle input changes through the lens of FGSM and high-dimensional geometry.2d ago
InSciforcebySciforceAdversarial Attacks Explained (And How to Defend ML Models Against Them)Simply put, the adversarial attack is a deceiving technique that is “fooling” machine learning models using a defective input. Adversarial…Sep 7, 2022
InAI ReporterbyRadha MittalOpenAI Boosts AI Security with Advanced Red Teaming Tactics 🔐In a world where AI evolves faster than your phone’s obsolescence rate, keeping these brainy bots on a leash is more important than ever…Nov 301
InTowards Data SciencebyMaham HaroonTo 1 or to 0: Pixel Attacks in Image ClassificationNavigating the Realm of Adversarial Machine LearningNov 23, 20232
