Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384302803> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W4384302803 abstract "Large-scale pre-trained models such as CodeBERT, GraphCodeBERT have earned widespread attention from both academia and industry. Attributed to the superior ability in code representation, they have been further applied in multiple downstream tasks such as clone detection, code search and code translation. However, it is also observed that these state-of-the-art pre-trained models are susceptible to adversarial attacks. The performance of these pre-trained models drops significantly with simple perturbations such as renaming variable names. This weakness may be inherited by their downstream models and thereby amplified at an unprecedented scale. To this end, we propose an approach namely ContraBERT that aims to improve the robustness of pre-trained models via contrastive learning. Specifically, we design nine kinds of simple and complex data augmentation operators on the programming language (PL) and natural language (NL) data to construct different variants. Furthermore, we continue to train the existing pre-trained models by masked language modeling (MLM) and contrastive pre-training task on the original samples with their augmented variants to enhance the robustness of the model. The extensive ex-periments demonstrate that ContraBERT can effectively improve the robustness of the existing pre-trained models. Further study also confirms that these robustness-enhanced models provide improvements as compared to original models over four popular downstream tasks." @default.
- W4384302803 created "2023-07-15" @default.
- W4384302803 creator A5017417068 @default.
- W4384302803 creator A5079237288 @default.
- W4384302803 creator A5079787466 @default.
- W4384302803 creator A5084396416 @default.
- W4384302803 date "2023-05-01" @default.
- W4384302803 modified "2023-10-17" @default.
- W4384302803 title "ContraBERT: Enhancing Code Pre-trained Models via Contrastive Learning" @default.
- W4384302803 cites W1980730196 @default.
- W4384302803 cites W1992114977 @default.
- W4384302803 cites W2064675550 @default.
- W4384302803 cites W2065053490 @default.
- W4384302803 cites W2282866165 @default.
- W4384302803 cites W2516621648 @default.
- W4384302803 cites W2794601162 @default.
- W4384302803 cites W2798991696 @default.
- W4384302803 cites W2962960733 @default.
- W4384302803 cites W2963935794 @default.
- W4384302803 cites W2964150020 @default.
- W4384302803 cites W2972082064 @default.
- W4384302803 cites W2979792666 @default.
- W4384302803 cites W2997451752 @default.
- W4384302803 cites W3014339000 @default.
- W4384302803 cites W3035524453 @default.
- W4384302803 cites W3098605233 @default.
- W4384302803 cites W3101228802 @default.
- W4384302803 cites W3108032709 @default.
- W4384302803 cites W3109966548 @default.
- W4384302803 cites W3121707215 @default.
- W4384302803 cites W3156636935 @default.
- W4384302803 cites W3159942024 @default.
- W4384302803 cites W3170092793 @default.
- W4384302803 cites W3174414731 @default.
- W4384302803 cites W3198685994 @default.
- W4384302803 cites W4220722393 @default.
- W4384302803 cites W4287889995 @default.
- W4384302803 cites W4313563520 @default.
- W4384302803 doi "https://doi.org/10.1109/icse48619.2023.00207" @default.
- W4384302803 hasPublicationYear "2023" @default.
- W4384302803 type Work @default.
- W4384302803 citedByCount "0" @default.
- W4384302803 crossrefType "proceedings-article" @default.
- W4384302803 hasAuthorship W4384302803A5017417068 @default.
- W4384302803 hasAuthorship W4384302803A5079237288 @default.
- W4384302803 hasAuthorship W4384302803A5079787466 @default.
- W4384302803 hasAuthorship W4384302803A5084396416 @default.
- W4384302803 hasBestOaLocation W43843028032 @default.
- W4384302803 hasConcept C104317684 @default.
- W4384302803 hasConcept C119857082 @default.
- W4384302803 hasConcept C137293760 @default.
- W4384302803 hasConcept C154945302 @default.
- W4384302803 hasConcept C185592680 @default.
- W4384302803 hasConcept C195324797 @default.
- W4384302803 hasConcept C203005215 @default.
- W4384302803 hasConcept C204321447 @default.
- W4384302803 hasConcept C41008148 @default.
- W4384302803 hasConcept C51632099 @default.
- W4384302803 hasConcept C55493867 @default.
- W4384302803 hasConcept C63479239 @default.
- W4384302803 hasConceptScore W4384302803C104317684 @default.
- W4384302803 hasConceptScore W4384302803C119857082 @default.
- W4384302803 hasConceptScore W4384302803C137293760 @default.
- W4384302803 hasConceptScore W4384302803C154945302 @default.
- W4384302803 hasConceptScore W4384302803C185592680 @default.
- W4384302803 hasConceptScore W4384302803C195324797 @default.
- W4384302803 hasConceptScore W4384302803C203005215 @default.
- W4384302803 hasConceptScore W4384302803C204321447 @default.
- W4384302803 hasConceptScore W4384302803C41008148 @default.
- W4384302803 hasConceptScore W4384302803C51632099 @default.
- W4384302803 hasConceptScore W4384302803C55493867 @default.
- W4384302803 hasConceptScore W4384302803C63479239 @default.
- W4384302803 hasLocation W43843028031 @default.
- W4384302803 hasLocation W43843028032 @default.
- W4384302803 hasOpenAccess W4384302803 @default.
- W4384302803 hasPrimaryLocation W43843028031 @default.
- W4384302803 hasRelatedWork W1569841287 @default.
- W4384302803 hasRelatedWork W1985007624 @default.
- W4384302803 hasRelatedWork W2176369193 @default.
- W4384302803 hasRelatedWork W2251559993 @default.
- W4384302803 hasRelatedWork W2747680751 @default.
- W4384302803 hasRelatedWork W2985347586 @default.
- W4384302803 hasRelatedWork W3107474891 @default.
- W4384302803 hasRelatedWork W3120880016 @default.
- W4384302803 hasRelatedWork W3198474835 @default.
- W4384302803 hasRelatedWork W4312189131 @default.
- W4384302803 isParatext "false" @default.
- W4384302803 isRetracted "false" @default.
- W4384302803 workType "article" @default.