Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311556220> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4311556220 abstract "This work addresses the problems of (a) designing utilization measurements of trained artificial intelligence (AI) models and (b) explaining how training data are encoded in AI models based on those measurements. The problems are motivated by the lack of explainability of AI models in security and safety critical applications, such as the use of AI models for classification of traffic signs in self-driving cars. We approach the problems by introducing theoretical underpinnings of AI model utilization measurement and understanding patterns in utilization-based class encodings of traffic signs at the level of computation graphs (AI models), subgraphs, and graph nodes. Conceptually, utilization is defined at each graph node (computation unit) of an AI model based on the number and distribution of unique outputs in the space of all possible outputs (tensor-states). In this work, utilization measurements are extracted from AI models, which include poisoned and clean AI models. In contrast to clean AI models, the poisoned AI models were trained with traffic sign images containing systematic, physically realizable, traffic sign modifications (i.e., triggers) to change a correct class label to another label in a presence of such a trigger. We analyze class encodings of such clean and poisoned AI models, and conclude with implications for trojan injection and detection." @default.
- W4311556220 created "2022-12-27" @default.
- W4311556220 creator A5003043661 @default.
- W4311556220 creator A5014638356 @default.
- W4311556220 creator A5015678162 @default.
- W4311556220 creator A5027650126 @default.
- W4311556220 creator A5033513737 @default.
- W4311556220 creator A5056299590 @default.
- W4311556220 creator A5071967448 @default.
- W4311556220 creator A5079607365 @default.
- W4311556220 date "2022-12-11" @default.
- W4311556220 modified "2023-09-27" @default.
- W4311556220 title "AI Model Utilization Measurements For Finding Class Encoding Patterns" @default.
- W4311556220 doi "https://doi.org/10.48550/arxiv.2212.06576" @default.
- W4311556220 hasPublicationYear "2022" @default.
- W4311556220 type Work @default.
- W4311556220 citedByCount "0" @default.
- W4311556220 crossrefType "posted-content" @default.
- W4311556220 hasAuthorship W4311556220A5003043661 @default.
- W4311556220 hasAuthorship W4311556220A5014638356 @default.
- W4311556220 hasAuthorship W4311556220A5015678162 @default.
- W4311556220 hasAuthorship W4311556220A5027650126 @default.
- W4311556220 hasAuthorship W4311556220A5033513737 @default.
- W4311556220 hasAuthorship W4311556220A5056299590 @default.
- W4311556220 hasAuthorship W4311556220A5071967448 @default.
- W4311556220 hasAuthorship W4311556220A5079607365 @default.
- W4311556220 hasBestOaLocation W43115562201 @default.
- W4311556220 hasConcept C11413529 @default.
- W4311556220 hasConcept C119857082 @default.
- W4311556220 hasConcept C132525143 @default.
- W4311556220 hasConcept C134306372 @default.
- W4311556220 hasConcept C139676723 @default.
- W4311556220 hasConcept C154945302 @default.
- W4311556220 hasConcept C2777212361 @default.
- W4311556220 hasConcept C2983860417 @default.
- W4311556220 hasConcept C33923547 @default.
- W4311556220 hasConcept C41008148 @default.
- W4311556220 hasConcept C45374587 @default.
- W4311556220 hasConcept C6528762 @default.
- W4311556220 hasConcept C80444323 @default.
- W4311556220 hasConceptScore W4311556220C11413529 @default.
- W4311556220 hasConceptScore W4311556220C119857082 @default.
- W4311556220 hasConceptScore W4311556220C132525143 @default.
- W4311556220 hasConceptScore W4311556220C134306372 @default.
- W4311556220 hasConceptScore W4311556220C139676723 @default.
- W4311556220 hasConceptScore W4311556220C154945302 @default.
- W4311556220 hasConceptScore W4311556220C2777212361 @default.
- W4311556220 hasConceptScore W4311556220C2983860417 @default.
- W4311556220 hasConceptScore W4311556220C33923547 @default.
- W4311556220 hasConceptScore W4311556220C41008148 @default.
- W4311556220 hasConceptScore W4311556220C45374587 @default.
- W4311556220 hasConceptScore W4311556220C6528762 @default.
- W4311556220 hasConceptScore W4311556220C80444323 @default.
- W4311556220 hasLocation W43115562201 @default.
- W4311556220 hasOpenAccess W4311556220 @default.
- W4311556220 hasPrimaryLocation W43115562201 @default.
- W4311556220 hasRelatedWork W2354062721 @default.
- W4311556220 hasRelatedWork W2908068254 @default.
- W4311556220 hasRelatedWork W2924149867 @default.
- W4311556220 hasRelatedWork W2961085424 @default.
- W4311556220 hasRelatedWork W3046775127 @default.
- W4311556220 hasRelatedWork W4205958290 @default.
- W4311556220 hasRelatedWork W4286629047 @default.
- W4311556220 hasRelatedWork W4306321456 @default.
- W4311556220 hasRelatedWork W4306674287 @default.
- W4311556220 hasRelatedWork W4224009465 @default.
- W4311556220 isParatext "false" @default.
- W4311556220 isRetracted "false" @default.
- W4311556220 workType "article" @default.