Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385571978> ?p ?o ?g. }
Showing items 1 to 55 of
55
with 100 items per page.
- W4385571978 abstract "Utilizing language models (LMs) without internal access is becoming an attractive paradigm in the field of NLP as many cutting-edge LMs are released through APIs and boast a massive scale.The de-facto method in this type of black-box scenario is known as prompting, which has shown progressive performance enhancements in situations where data labels are scarce or unavailable.Despite their efficacy, they still fall short in comparison to fully supervised counterparts and are generally brittle to slight modifications.In this paper, we propose Clustering-enhanced Linear Discriminative Analysis (CELDA), a novel approach that improves the text classification accuracy with a very weak-supervision signal (i.e., name of the labels).Our framework draws a precise decision boundary without accessing weights or gradients of the LM model or data labels.The core ideas of CELDA are twofold:(1) extracting a refined pseudo-labeled dataset from an unlabeled dataset, and (2) training a lightweight and robust model on the top of LM, which learns an accurate decision boundary from an extracted noisy dataset.Throughout in-depth investigations on various datasets, we demonstrated that CELDA reaches new state-of-the-art in weakly-supervised text classification and narrows the gap with a fully-supervised model.Additionally, our proposed methodology can be applied universally to any LM and has the potential to scale to larger models, making it a more viable option for utilizing large LMs." @default.
- W4385571978 created "2023-08-05" @default.
- W4385571978 creator A5015505425 @default.
- W4385571978 creator A5062457999 @default.
- W4385571978 creator A5075095366 @default.
- W4385571978 date "2023-01-01" @default.
- W4385571978 modified "2023-09-23" @default.
- W4385571978 title "CELDA: Leveraging Black-box Language Model as Enhanced Classifier without Labels" @default.
- W4385571978 doi "https://doi.org/10.18653/v1/2023.acl-long.239" @default.
- W4385571978 hasPublicationYear "2023" @default.
- W4385571978 type Work @default.
- W4385571978 citedByCount "0" @default.
- W4385571978 crossrefType "proceedings-article" @default.
- W4385571978 hasAuthorship W4385571978A5015505425 @default.
- W4385571978 hasAuthorship W4385571978A5062457999 @default.
- W4385571978 hasAuthorship W4385571978A5075095366 @default.
- W4385571978 hasBestOaLocation W43855719781 @default.
- W4385571978 hasConcept C119857082 @default.
- W4385571978 hasConcept C137293760 @default.
- W4385571978 hasConcept C153180895 @default.
- W4385571978 hasConcept C154945302 @default.
- W4385571978 hasConcept C204321447 @default.
- W4385571978 hasConcept C41008148 @default.
- W4385571978 hasConcept C42023084 @default.
- W4385571978 hasConcept C73555534 @default.
- W4385571978 hasConcept C774472 @default.
- W4385571978 hasConcept C95623464 @default.
- W4385571978 hasConcept C97931131 @default.
- W4385571978 hasConceptScore W4385571978C119857082 @default.
- W4385571978 hasConceptScore W4385571978C137293760 @default.
- W4385571978 hasConceptScore W4385571978C153180895 @default.
- W4385571978 hasConceptScore W4385571978C154945302 @default.
- W4385571978 hasConceptScore W4385571978C204321447 @default.
- W4385571978 hasConceptScore W4385571978C41008148 @default.
- W4385571978 hasConceptScore W4385571978C42023084 @default.
- W4385571978 hasConceptScore W4385571978C73555534 @default.
- W4385571978 hasConceptScore W4385571978C774472 @default.
- W4385571978 hasConceptScore W4385571978C95623464 @default.
- W4385571978 hasConceptScore W4385571978C97931131 @default.
- W4385571978 hasLocation W43855719781 @default.
- W4385571978 hasOpenAccess W4385571978 @default.
- W4385571978 hasPrimaryLocation W43855719781 @default.
- W4385571978 hasRelatedWork W1972656095 @default.
- W4385571978 hasRelatedWork W2024160000 @default.
- W4385571978 hasRelatedWork W2061273563 @default.
- W4385571978 hasRelatedWork W2285052147 @default.
- W4385571978 hasRelatedWork W2729514902 @default.
- W4385571978 hasRelatedWork W2743258233 @default.
- W4385571978 hasRelatedWork W2773500201 @default.
- W4385571978 hasRelatedWork W2970216048 @default.
- W4385571978 hasRelatedWork W2998168123 @default.
- W4385571978 hasRelatedWork W4287995534 @default.
- W4385571978 isParatext "false" @default.
- W4385571978 isRetracted "false" @default.
- W4385571978 workType "article" @default.