Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313441748> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W4313441748 abstract "We develop and study new adversarial perturbations that enable an attacker to gain control over decisions in generic Artificial Intelligence (AI) systems including deep learning neural networks. In contrast to adversarial data modification, the attack mechanism we consider here involves alterations to the AI system itself. Such a stealth attack could be conducted by a mischievous, corrupt or disgruntled member of a software development team. It could also be made by those wishing to exploit a ``democratization of AI'' agenda, where network architectures and trained parameter sets are shared publicly. We develop a range of new implementable attack strategies with accompanying analysis, showing that with high probability a stealth attack can be made transparent, in the sense that system performance is unchanged on a fixed validation set which is unknown to the attacker, while evoking any desired output on a trigger input of interest. The attacker only needs to have estimates of the size of the validation set and the spread of the AI's relevant latent space. In the case of deep learning neural networks, we show that a one neuron attack is possible - a modification to the weights and bias associated with a single neuron - revealing a vulnerability arising from over-parameterization. We illustrate these concepts using state of the art architectures on two standard image data sets. Guided by the theory and computational results, we also propose strategies to guard against stealth attacks." @default.
- W4313441748 created "2023-01-06" @default.
- W4313441748 creator A5004339750 @default.
- W4313441748 creator A5052143104 @default.
- W4313441748 creator A5058069510 @default.
- W4313441748 creator A5058486589 @default.
- W4313441748 creator A5088165793 @default.
- W4313441748 date "2021-06-26" @default.
- W4313441748 modified "2023-10-16" @default.
- W4313441748 title "The Feasibility and Inevitability of Stealth Attacks" @default.
- W4313441748 doi "https://doi.org/10.48550/arxiv.2106.13997" @default.
- W4313441748 hasPublicationYear "2021" @default.
- W4313441748 type Work @default.
- W4313441748 citedByCount "1" @default.
- W4313441748 countsByYear W43134417482023 @default.
- W4313441748 crossrefType "posted-content" @default.
- W4313441748 hasAuthorship W4313441748A5004339750 @default.
- W4313441748 hasAuthorship W4313441748A5052143104 @default.
- W4313441748 hasAuthorship W4313441748A5058069510 @default.
- W4313441748 hasAuthorship W4313441748A5058486589 @default.
- W4313441748 hasAuthorship W4313441748A5088165793 @default.
- W4313441748 hasBestOaLocation W43134417481 @default.
- W4313441748 hasConcept C108583219 @default.
- W4313441748 hasConcept C119857082 @default.
- W4313441748 hasConcept C141141315 @default.
- W4313441748 hasConcept C154945302 @default.
- W4313441748 hasConcept C165696696 @default.
- W4313441748 hasConcept C177264268 @default.
- W4313441748 hasConcept C199360897 @default.
- W4313441748 hasConcept C37736160 @default.
- W4313441748 hasConcept C38652104 @default.
- W4313441748 hasConcept C41008148 @default.
- W4313441748 hasConcept C50644808 @default.
- W4313441748 hasConcept C80444323 @default.
- W4313441748 hasConcept C95713431 @default.
- W4313441748 hasConceptScore W4313441748C108583219 @default.
- W4313441748 hasConceptScore W4313441748C119857082 @default.
- W4313441748 hasConceptScore W4313441748C141141315 @default.
- W4313441748 hasConceptScore W4313441748C154945302 @default.
- W4313441748 hasConceptScore W4313441748C165696696 @default.
- W4313441748 hasConceptScore W4313441748C177264268 @default.
- W4313441748 hasConceptScore W4313441748C199360897 @default.
- W4313441748 hasConceptScore W4313441748C37736160 @default.
- W4313441748 hasConceptScore W4313441748C38652104 @default.
- W4313441748 hasConceptScore W4313441748C41008148 @default.
- W4313441748 hasConceptScore W4313441748C50644808 @default.
- W4313441748 hasConceptScore W4313441748C80444323 @default.
- W4313441748 hasConceptScore W4313441748C95713431 @default.
- W4313441748 hasLocation W43134417481 @default.
- W4313441748 hasOpenAccess W4313441748 @default.
- W4313441748 hasPrimaryLocation W43134417481 @default.
- W4313441748 hasRelatedWork W2042616262 @default.
- W4313441748 hasRelatedWork W286553814 @default.
- W4313441748 hasRelatedWork W3014300295 @default.
- W4313441748 hasRelatedWork W3122267592 @default.
- W4313441748 hasRelatedWork W3124408655 @default.
- W4313441748 hasRelatedWork W4223943233 @default.
- W4313441748 hasRelatedWork W4225161397 @default.
- W4313441748 hasRelatedWork W4297785512 @default.
- W4313441748 hasRelatedWork W4309045103 @default.
- W4313441748 hasRelatedWork W4312200629 @default.
- W4313441748 isParatext "false" @default.
- W4313441748 isRetracted "false" @default.
- W4313441748 workType "article" @default.