Matches in SemOpenAlex for { <https://semopenalex.org/work/W3124249192> ?p ?o ?g. }
- W3124249192 abstract "Growing at a fast pace, modern autonomous systems will soon be deployed at scale, opening up the possibility for cooperative multi-agent systems. Sharing information and distributing workloads allow autonomous agents to better perform tasks and increase computation efficiency. However, shared information can be modified to execute adversarial attacks on deep learning models that are widely employed in modern systems. Thus, we aim to study the robustness of such systems and focus on exploring adversarial attacks in a novel multi-agent setting where communication is done through sharing learned intermediate representations of neural networks. We observe that an indistinguishable adversarial message can severely degrade performance, but becomes weaker as the number of benign agents increases. Furthermore, we show that black-box transfer attacks are more difficult in this setting when compared to directly perturbing the inputs, as it is necessary to align the distribution of learned representations with domain adaptation. Our work studies robustness at the neural network level to contribute an additional layer of fault tolerance to modern security protocols for more secure multi-agent systems." @default.
- W3124249192 created "2021-02-01" @default.
- W3124249192 creator A5000380827 @default.
- W3124249192 creator A5010890057 @default.
- W3124249192 creator A5017563787 @default.
- W3124249192 creator A5026746739 @default.
- W3124249192 creator A5041739049 @default.
- W3124249192 creator A5058557954 @default.
- W3124249192 date "2021-01-16" @default.
- W3124249192 modified "2023-09-27" @default.
- W3124249192 title "Adversarial Attacks On Multi-Agent Communication" @default.
- W3124249192 cites W1217034487 @default.
- W3124249192 cites W1826987073 @default.
- W3124249192 cites W1901129140 @default.
- W3124249192 cites W2015252327 @default.
- W3124249192 cites W2028397517 @default.
- W3124249192 cites W2092747249 @default.
- W3124249192 cites W2098611419 @default.
- W3124249192 cites W2127133497 @default.
- W3124249192 cites W2150595850 @default.
- W3124249192 cites W2166130341 @default.
- W3124249192 cites W2190691619 @default.
- W3124249192 cites W2535838896 @default.
- W3124249192 cites W2593768305 @default.
- W3124249192 cites W2603766943 @default.
- W3124249192 cites W2604505099 @default.
- W3124249192 cites W2746600820 @default.
- W3124249192 cites W2760600531 @default.
- W3124249192 cites W2806506455 @default.
- W3124249192 cites W2898979134 @default.
- W3124249192 cites W2906208681 @default.
- W3124249192 cites W2949103145 @default.
- W3124249192 cites W2949505878 @default.
- W3124249192 cites W2962847335 @default.
- W3124249192 cites W2962917037 @default.
- W3124249192 cites W2963070423 @default.
- W3124249192 cites W2963083752 @default.
- W3124249192 cites W2963207607 @default.
- W3124249192 cites W2963351448 @default.
- W3124249192 cites W2963422767 @default.
- W3124249192 cites W2963834268 @default.
- W3124249192 cites W2963836885 @default.
- W3124249192 cites W2963857521 @default.
- W3124249192 cites W2963981733 @default.
- W3124249192 cites W2964116600 @default.
- W3124249192 cites W2964153729 @default.
- W3124249192 cites W2964253222 @default.
- W3124249192 cites W2964346747 @default.
- W3124249192 cites W2971126145 @default.
- W3124249192 cites W2982109374 @default.
- W3124249192 cites W2984699060 @default.
- W3124249192 cites W2995164118 @default.
- W3124249192 cites W2998277219 @default.
- W3124249192 cites W3006647218 @default.
- W3124249192 cites W3013577831 @default.
- W3124249192 cites W3015625436 @default.
- W3124249192 cites W3034537217 @default.
- W3124249192 cites W3034716600 @default.
- W3124249192 cites W3038028469 @default.
- W3124249192 cites W3048715803 @default.
- W3124249192 cites W3048723201 @default.
- W3124249192 cites W3098881644 @default.
- W3124249192 cites W3109991383 @default.
- W3124249192 hasPublicationYear "2021" @default.
- W3124249192 type Work @default.
- W3124249192 sameAs 3124249192 @default.
- W3124249192 citedByCount "1" @default.
- W3124249192 countsByYear W31242491922021 @default.
- W3124249192 crossrefType "posted-content" @default.
- W3124249192 hasAuthorship W3124249192A5000380827 @default.
- W3124249192 hasAuthorship W3124249192A5010890057 @default.
- W3124249192 hasAuthorship W3124249192A5017563787 @default.
- W3124249192 hasAuthorship W3124249192A5026746739 @default.
- W3124249192 hasAuthorship W3124249192A5041739049 @default.
- W3124249192 hasAuthorship W3124249192A5058557954 @default.
- W3124249192 hasBestOaLocation W31242491921 @default.
- W3124249192 hasConcept C104317684 @default.
- W3124249192 hasConcept C119857082 @default.
- W3124249192 hasConcept C120314980 @default.
- W3124249192 hasConcept C120665830 @default.
- W3124249192 hasConcept C121332964 @default.
- W3124249192 hasConcept C154945302 @default.
- W3124249192 hasConcept C185592680 @default.
- W3124249192 hasConcept C192209626 @default.
- W3124249192 hasConcept C2984842247 @default.
- W3124249192 hasConcept C37736160 @default.
- W3124249192 hasConcept C38652104 @default.
- W3124249192 hasConcept C41008148 @default.
- W3124249192 hasConcept C50644808 @default.
- W3124249192 hasConcept C55493867 @default.
- W3124249192 hasConcept C63479239 @default.
- W3124249192 hasConcept C63540848 @default.
- W3124249192 hasConceptScore W3124249192C104317684 @default.
- W3124249192 hasConceptScore W3124249192C119857082 @default.
- W3124249192 hasConceptScore W3124249192C120314980 @default.
- W3124249192 hasConceptScore W3124249192C120665830 @default.
- W3124249192 hasConceptScore W3124249192C121332964 @default.
- W3124249192 hasConceptScore W3124249192C154945302 @default.
- W3124249192 hasConceptScore W3124249192C185592680 @default.
- W3124249192 hasConceptScore W3124249192C192209626 @default.