Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287890478> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4287890478 abstract "Deep learning based systems are susceptible to adversarial attacks, where a small, imperceptible change at the input alters the model prediction. However, to date the majority of the approaches to detect these attacks have been designed for image processing systems. Many popular image adversarial detection approaches are able to identify adversarial examples from embedding feature spaces, whilst in the NLP domain existing state of the art detection approaches solely focus on input text features, without consideration of model embedding spaces. This work examines what differences result when porting these image designed strategies to Natural Language Processing (NLP) tasks - these detectors are found to not port over well. This is expected as NLP systems have a very different form of input: discrete and sequential in nature, rather than the continuous and fixed size inputs for images. As an equivalent model-focused NLP detection approach, this work proposes a simple sentence-embedding “residue” based detector to identify adversarial examples. On many tasks, it out-performs ported image domain detectors and recent state of the art NLP specific detectors." @default.
- W4287890478 created "2022-07-26" @default.
- W4287890478 creator A5024132945 @default.
- W4287890478 creator A5050766679 @default.
- W4287890478 date "2022-01-01" @default.
- W4287890478 modified "2023-10-18" @default.
- W4287890478 title "Residue-Based Natural Language Adversarial Attack Detection" @default.
- W4287890478 doi "https://doi.org/10.18653/v1/2022.naacl-main.281" @default.
- W4287890478 hasPublicationYear "2022" @default.
- W4287890478 type Work @default.
- W4287890478 citedByCount "0" @default.
- W4287890478 crossrefType "proceedings-article" @default.
- W4287890478 hasAuthorship W4287890478A5024132945 @default.
- W4287890478 hasAuthorship W4287890478A5050766679 @default.
- W4287890478 hasBestOaLocation W42878904781 @default.
- W4287890478 hasConcept C106251023 @default.
- W4287890478 hasConcept C108583219 @default.
- W4287890478 hasConcept C115961682 @default.
- W4287890478 hasConcept C119857082 @default.
- W4287890478 hasConcept C120665830 @default.
- W4287890478 hasConcept C121332964 @default.
- W4287890478 hasConcept C153180895 @default.
- W4287890478 hasConcept C154945302 @default.
- W4287890478 hasConcept C192209626 @default.
- W4287890478 hasConcept C195324797 @default.
- W4287890478 hasConcept C199360897 @default.
- W4287890478 hasConcept C204321447 @default.
- W4287890478 hasConcept C2777530160 @default.
- W4287890478 hasConcept C2777904410 @default.
- W4287890478 hasConcept C37736160 @default.
- W4287890478 hasConcept C41008148 @default.
- W4287890478 hasConcept C41608201 @default.
- W4287890478 hasConcept C76155785 @default.
- W4287890478 hasConcept C94915269 @default.
- W4287890478 hasConceptScore W4287890478C106251023 @default.
- W4287890478 hasConceptScore W4287890478C108583219 @default.
- W4287890478 hasConceptScore W4287890478C115961682 @default.
- W4287890478 hasConceptScore W4287890478C119857082 @default.
- W4287890478 hasConceptScore W4287890478C120665830 @default.
- W4287890478 hasConceptScore W4287890478C121332964 @default.
- W4287890478 hasConceptScore W4287890478C153180895 @default.
- W4287890478 hasConceptScore W4287890478C154945302 @default.
- W4287890478 hasConceptScore W4287890478C192209626 @default.
- W4287890478 hasConceptScore W4287890478C195324797 @default.
- W4287890478 hasConceptScore W4287890478C199360897 @default.
- W4287890478 hasConceptScore W4287890478C204321447 @default.
- W4287890478 hasConceptScore W4287890478C2777530160 @default.
- W4287890478 hasConceptScore W4287890478C2777904410 @default.
- W4287890478 hasConceptScore W4287890478C37736160 @default.
- W4287890478 hasConceptScore W4287890478C41008148 @default.
- W4287890478 hasConceptScore W4287890478C41608201 @default.
- W4287890478 hasConceptScore W4287890478C76155785 @default.
- W4287890478 hasConceptScore W4287890478C94915269 @default.
- W4287890478 hasLocation W42878904781 @default.
- W4287890478 hasLocation W42878904782 @default.
- W4287890478 hasLocation W42878904783 @default.
- W4287890478 hasOpenAccess W4287890478 @default.
- W4287890478 hasPrimaryLocation W42878904781 @default.
- W4287890478 hasRelatedWork W159132833 @default.
- W4287890478 hasRelatedWork W3014300295 @default.
- W4287890478 hasRelatedWork W3164822677 @default.
- W4287890478 hasRelatedWork W4223943233 @default.
- W4287890478 hasRelatedWork W4225161397 @default.
- W4287890478 hasRelatedWork W4312200629 @default.
- W4287890478 hasRelatedWork W4360585206 @default.
- W4287890478 hasRelatedWork W4364306694 @default.
- W4287890478 hasRelatedWork W4380075502 @default.
- W4287890478 hasRelatedWork W4380086463 @default.
- W4287890478 isParatext "false" @default.
- W4287890478 isRetracted "false" @default.
- W4287890478 workType "article" @default.