Matches in SemOpenAlex for { <https://semopenalex.org/work/W3044324512> ?p ?o ?g. }
- W3044324512 abstract "While there has been substantial research using adversarial attacks to analyze NLP models, each attack is implemented in its own code repository. It remains challenging to develop NLP attacks and utilize them to improve model performance. This paper introduces TextAttack, a Python framework for adversarial attacks, data augmentation, and adversarial training in NLP. TextAttack builds attacks from four components: a goal function, a set of constraints, a transformation, and a search method. TextAttack's modular design enables researchers to easily construct attacks from combinations of novel and existing components. TextAttack provides implementations of 16 adversarial attacks from the literature and supports a variety of models and datasets, including BERT and other transformers, and all GLUE tasks. TextAttack also includes data augmentation and adversarial training modules for using components of adversarial attacks to improve model accuracy and robustness. TextAttack is democratizing NLP: anyone can try data augmentation and adversarial training on any model or dataset, with just a few lines of code. Code and tutorials are available at https://github.com/QData/TextAttack." @default.
- W3044324512 created "2020-07-29" @default.
- W3044324512 creator A5005724403 @default.
- W3044324512 creator A5012455357 @default.
- W3044324512 creator A5014105687 @default.
- W3044324512 creator A5029164534 @default.
- W3044324512 creator A5071517081 @default.
- W3044324512 creator A5078952987 @default.
- W3044324512 date "2020-04-29" @default.
- W3044324512 modified "2023-09-23" @default.
- W3044324512 title "TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP" @default.
- W3044324512 cites W1591840823 @default.
- W3044324512 cites W2101105183 @default.
- W3044324512 cites W2102381086 @default.
- W3044324512 cites W2250342921 @default.
- W3044324512 cites W2259472270 @default.
- W3044324512 cites W2289952147 @default.
- W3044324512 cites W2543580944 @default.
- W3044324512 cites W2785699986 @default.
- W3044324512 cites W2791941932 @default.
- W3044324512 cites W2794557536 @default.
- W3044324512 cites W2798966449 @default.
- W3044324512 cites W2799194071 @default.
- W3044324512 cites W2810611310 @default.
- W3044324512 cites W2898268570 @default.
- W3044324512 cites W2922293812 @default.
- W3044324512 cites W2947415936 @default.
- W3044324512 cites W2949128310 @default.
- W3044324512 cites W2950726992 @default.
- W3044324512 cites W2962818281 @default.
- W3044324512 cites W2963283805 @default.
- W3044324512 cites W2963859254 @default.
- W3044324512 cites W2963918774 @default.
- W3044324512 cites W2963969878 @default.
- W3044324512 cites W2964153729 @default.
- W3044324512 cites W2965595599 @default.
- W3044324512 cites W2970078867 @default.
- W3044324512 cites W2970863760 @default.
- W3044324512 cites W2971296908 @default.
- W3044324512 cites W2972619028 @default.
- W3044324512 cites W2996403597 @default.
- W3044324512 cites W3014153848 @default.
- W3044324512 cites W3018458867 @default.
- W3044324512 cites W3035736465 @default.
- W3044324512 cites W2886991642 @default.
- W3044324512 doi "https://doi.org/10.48550/arxiv.2005.05909" @default.
- W3044324512 hasPublicationYear "2020" @default.
- W3044324512 type Work @default.
- W3044324512 sameAs 3044324512 @default.
- W3044324512 citedByCount "24" @default.
- W3044324512 countsByYear W30443245122020 @default.
- W3044324512 countsByYear W30443245122021 @default.
- W3044324512 crossrefType "posted-content" @default.
- W3044324512 hasAuthorship W3044324512A5005724403 @default.
- W3044324512 hasAuthorship W3044324512A5012455357 @default.
- W3044324512 hasAuthorship W3044324512A5014105687 @default.
- W3044324512 hasAuthorship W3044324512A5029164534 @default.
- W3044324512 hasAuthorship W3044324512A5071517081 @default.
- W3044324512 hasAuthorship W3044324512A5078952987 @default.
- W3044324512 hasBestOaLocation W30443245121 @default.
- W3044324512 hasConcept C101468663 @default.
- W3044324512 hasConcept C104317684 @default.
- W3044324512 hasConcept C119857082 @default.
- W3044324512 hasConcept C121332964 @default.
- W3044324512 hasConcept C154945302 @default.
- W3044324512 hasConcept C165801399 @default.
- W3044324512 hasConcept C177264268 @default.
- W3044324512 hasConcept C185592680 @default.
- W3044324512 hasConcept C199360897 @default.
- W3044324512 hasConcept C199519371 @default.
- W3044324512 hasConcept C204321447 @default.
- W3044324512 hasConcept C26713055 @default.
- W3044324512 hasConcept C2776760102 @default.
- W3044324512 hasConcept C2777904410 @default.
- W3044324512 hasConcept C37736160 @default.
- W3044324512 hasConcept C41008148 @default.
- W3044324512 hasConcept C51632099 @default.
- W3044324512 hasConcept C519991488 @default.
- W3044324512 hasConcept C55493867 @default.
- W3044324512 hasConcept C62520636 @default.
- W3044324512 hasConcept C63479239 @default.
- W3044324512 hasConcept C66322947 @default.
- W3044324512 hasConceptScore W3044324512C101468663 @default.
- W3044324512 hasConceptScore W3044324512C104317684 @default.
- W3044324512 hasConceptScore W3044324512C119857082 @default.
- W3044324512 hasConceptScore W3044324512C121332964 @default.
- W3044324512 hasConceptScore W3044324512C154945302 @default.
- W3044324512 hasConceptScore W3044324512C165801399 @default.
- W3044324512 hasConceptScore W3044324512C177264268 @default.
- W3044324512 hasConceptScore W3044324512C185592680 @default.
- W3044324512 hasConceptScore W3044324512C199360897 @default.
- W3044324512 hasConceptScore W3044324512C199519371 @default.
- W3044324512 hasConceptScore W3044324512C204321447 @default.
- W3044324512 hasConceptScore W3044324512C26713055 @default.
- W3044324512 hasConceptScore W3044324512C2776760102 @default.
- W3044324512 hasConceptScore W3044324512C2777904410 @default.
- W3044324512 hasConceptScore W3044324512C37736160 @default.
- W3044324512 hasConceptScore W3044324512C41008148 @default.
- W3044324512 hasConceptScore W3044324512C51632099 @default.
- W3044324512 hasConceptScore W3044324512C519991488 @default.