FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3204191745> ?p ?o ?g. }

• W3204191745 abstract "Pre-trained models are widely used in fine-tuning downstream tasks with linear classifiers optimized by the cross-entropy loss, which might face robustness and stability problems. These problems can be improved by learning representations that focus on similarities in the same class and contradictions in different classes when making predictions. In this paper, we utilize the K-Nearest Neighbors Classifier in pre-trained model fine-tuning. For this KNN classifier, we introduce a supervised momentum contrastive learning framework to learn the clustered representations of the supervised downstream tasks. Extensive experiments on text classification tasks and robustness tests show that by incorporating KNNs with the traditional fine-tuning process, we can obtain significant improvements on the clean accuracy in both rich-source and few-shot settings and can improve the robustness against adversarial attacks. footnote{all codes is available at this https URL}" @default.
• W3204191745 created "2021-10-11" @default.
• W3204191745 creator A5010089116 @default.
• W3204191745 creator A5044665993 @default.
• W3204191745 creator A5086682667 @default.
• W3204191745 creator A5088256860 @default.
• W3204191745 creator A5088834359 @default.
• W3204191745 date "2021-10-06" @default.
• W3204191745 modified "2023-09-27" @default.
• W3204191745 title "KNN-BERT: Fine-Tuning Pre-Trained Models with KNN Classifier" @default.
• W3204191745 cites W131533222 @default.
• W3204191745 cites W1498436455 @default.
• W3204191745 cites W1866072925 @default.
• W3204191745 cites W2097732278 @default.
• W3204191745 cites W2106053110 @default.
• W3204191745 cites W2113459411 @default.
• W3204191745 cites W2138621090 @default.
• W3204191745 cites W2152790380 @default.
• W3204191745 cites W2194775991 @default.
• W3204191745 cites W2251939518 @default.
• W3204191745 cites W2601450892 @default.
• W3204191745 cites W2769112066 @default.
• W3204191745 cites W2793165286 @default.
• W3204191745 cites W2798991696 @default.
• W3204191745 cites W2842511635 @default.
• W3204191745 cites W2913881544 @default.
• W3204191745 cites W2951873722 @default.
• W3204191745 cites W2963012544 @default.
• W3204191745 cites W2963207607 @default.
• W3204191745 cites W2963341956 @default.
• W3204191745 cites W2963403868 @default.
• W3204191745 cites W2963656735 @default.
• W3204191745 cites W2963748441 @default.
• W3204191745 cites W2963759070 @default.
• W3204191745 cites W2963846996 @default.
• W3204191745 cites W2963857521 @default.
• W3204191745 cites W2964420626 @default.
• W3204191745 cites W2965373594 @default.
• W3204191745 cites W2965595599 @default.
• W3204191745 cites W2970597249 @default.
• W3204191745 cites W2970941190 @default.
• W3204191745 cites W2995154514 @default.
• W3204191745 cites W2996428491 @default.
• W3204191745 cites W3005700362 @default.
• W3204191745 cites W3034709122 @default.
• W3204191745 cites W3034978746 @default.
• W3204191745 cites W3035058308 @default.
• W3204191745 cites W3035524453 @default.
• W3204191745 cites W3099782249 @default.
• W3204191745 cites W3100345210 @default.
• W3204191745 cites W3102839769 @default.
• W3204191745 cites W3105604018 @default.
• W3204191745 cites W3122924117 @default.
• W3204191745 cites W2525127255 @default.
• W3204191745 hasPublicationYear "2021" @default.
• W3204191745 type Work @default.
• W3204191745 sameAs 3204191745 @default.
• W3204191745 citedByCount "0" @default.
• W3204191745 crossrefType "posted-content" @default.
• W3204191745 hasAuthorship W3204191745A5010089116 @default.
• W3204191745 hasAuthorship W3204191745A5044665993 @default.
• W3204191745 hasAuthorship W3204191745A5086682667 @default.
• W3204191745 hasAuthorship W3204191745A5088256860 @default.
• W3204191745 hasAuthorship W3204191745A5088834359 @default.
• W3204191745 hasConcept C104317684 @default.
• W3204191745 hasConcept C119857082 @default.
• W3204191745 hasConcept C121332964 @default.
• W3204191745 hasConcept C139532973 @default.
• W3204191745 hasConcept C153180895 @default.
• W3204191745 hasConcept C154945302 @default.
• W3204191745 hasConcept C157524613 @default.
• W3204191745 hasConcept C185592680 @default.
• W3204191745 hasConcept C41008148 @default.
• W3204191745 hasConcept C55493867 @default.
• W3204191745 hasConcept C62520636 @default.
• W3204191745 hasConcept C63479239 @default.
• W3204191745 hasConcept C95623464 @default.
• W3204191745 hasConceptScore W3204191745C104317684 @default.
• W3204191745 hasConceptScore W3204191745C119857082 @default.
• W3204191745 hasConceptScore W3204191745C121332964 @default.
• W3204191745 hasConceptScore W3204191745C139532973 @default.
• W3204191745 hasConceptScore W3204191745C153180895 @default.
• W3204191745 hasConceptScore W3204191745C154945302 @default.
• W3204191745 hasConceptScore W3204191745C157524613 @default.
• W3204191745 hasConceptScore W3204191745C185592680 @default.
• W3204191745 hasConceptScore W3204191745C41008148 @default.
• W3204191745 hasConceptScore W3204191745C55493867 @default.
• W3204191745 hasConceptScore W3204191745C62520636 @default.
• W3204191745 hasConceptScore W3204191745C63479239 @default.
• W3204191745 hasConceptScore W3204191745C95623464 @default.
• W3204191745 hasLocation W32041917451 @default.
• W3204191745 hasOpenAccess W3204191745 @default.
• W3204191745 hasPrimaryLocation W32041917451 @default.
• W3204191745 hasRelatedWork W103332046 @default.
• W3204191745 hasRelatedWork W1495850371 @default.
• W3204191745 hasRelatedWork W1585680390 @default.
• W3204191745 hasRelatedWork W1970077680 @default.
• W3204191745 hasRelatedWork W1984630326 @default.
• W3204191745 hasRelatedWork W1989790332 @default.
• W3204191745 hasRelatedWork W2013840542 @default.

• next