Matches in SemOpenAlex for { <https://semopenalex.org/work/W3085177480> ?p ?o ?g. }
- W3085177480 abstract "When scaled to hundreds of billions of parameters, pretrained language models such as GPT-3 (Brown et al., 2020) achieve remarkable few-shot performance. However, enormous amounts of compute are required for training and applying such big models, resulting in a large carbon footprint and making it difficult for researchers and practitioners to use them. We show that performance similar to GPT-3 can be obtained with language models that are much greener in that their parameter count is several orders of magnitude smaller. This is achieved by converting textual inputs into cloze questions that contain a task description, combined with gradient-based optimization; exploiting unlabeled data gives further improvements. We identify key factors required for successful natural language understanding with small language models." @default.
- W3085177480 created "2020-09-21" @default.
- W3085177480 creator A5071144367 @default.
- W3085177480 creator A5076269589 @default.
- W3085177480 date "2020-09-15" @default.
- W3085177480 modified "2023-09-27" @default.
- W3085177480 title "It's Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners" @default.
- W3085177480 cites W1489949474 @default.
- W3085177480 cites W1599016936 @default.
- W3085177480 cites W1724438581 @default.
- W3085177480 cites W1821462560 @default.
- W3085177480 cites W2111316763 @default.
- W3085177480 cites W2145755360 @default.
- W3085177480 cites W2804897457 @default.
- W3085177480 cites W2805206884 @default.
- W3085177480 cites W2898662126 @default.
- W3085177480 cites W2927746189 @default.
- W3085177480 cites W2956105246 @default.
- W3085177480 cites W2962847482 @default.
- W3085177480 cites W2963341956 @default.
- W3085177480 cites W2963674932 @default.
- W3085177480 cites W2963846996 @default.
- W3085177480 cites W2964299589 @default.
- W3085177480 cites W2965373594 @default.
- W3085177480 cites W2970597249 @default.
- W3085177480 cites W2978017171 @default.
- W3085177480 cites W2981852735 @default.
- W3085177480 cites W2982399380 @default.
- W3085177480 cites W2991382858 @default.
- W3085177480 cites W2995335514 @default.
- W3085177480 cites W2998554035 @default.
- W3085177480 cites W3004346089 @default.
- W3085177480 cites W3005700362 @default.
- W3085177480 cites W3015468748 @default.
- W3085177480 cites W3026404337 @default.
- W3085177480 cites W3029040966 @default.
- W3085177480 cites W3030163527 @default.
- W3085177480 cites W3111372685 @default.
- W3085177480 cites W3153427360 @default.
- W3085177480 cites W806995027 @default.
- W3085177480 cites W95183648 @default.
- W3085177480 hasPublicationYear "2020" @default.
- W3085177480 type Work @default.
- W3085177480 sameAs 3085177480 @default.
- W3085177480 citedByCount "45" @default.
- W3085177480 countsByYear W30851774802020 @default.
- W3085177480 countsByYear W30851774802021 @default.
- W3085177480 countsByYear W30851774802022 @default.
- W3085177480 crossrefType "posted-content" @default.
- W3085177480 hasAuthorship W3085177480A5071144367 @default.
- W3085177480 hasAuthorship W3085177480A5076269589 @default.
- W3085177480 hasConcept C127413603 @default.
- W3085177480 hasConcept C132943942 @default.
- W3085177480 hasConcept C137293760 @default.
- W3085177480 hasConcept C154945302 @default.
- W3085177480 hasConcept C166957645 @default.
- W3085177480 hasConcept C178790620 @default.
- W3085177480 hasConcept C185592680 @default.
- W3085177480 hasConcept C18903297 @default.
- W3085177480 hasConcept C201995342 @default.
- W3085177480 hasConcept C204321447 @default.
- W3085177480 hasConcept C205649164 @default.
- W3085177480 hasConcept C26517878 @default.
- W3085177480 hasConcept C2778344882 @default.
- W3085177480 hasConcept C2780451532 @default.
- W3085177480 hasConcept C2780936489 @default.
- W3085177480 hasConcept C38652104 @default.
- W3085177480 hasConcept C41008148 @default.
- W3085177480 hasConcept C47737302 @default.
- W3085177480 hasConcept C86803240 @default.
- W3085177480 hasConceptScore W3085177480C127413603 @default.
- W3085177480 hasConceptScore W3085177480C132943942 @default.
- W3085177480 hasConceptScore W3085177480C137293760 @default.
- W3085177480 hasConceptScore W3085177480C154945302 @default.
- W3085177480 hasConceptScore W3085177480C166957645 @default.
- W3085177480 hasConceptScore W3085177480C178790620 @default.
- W3085177480 hasConceptScore W3085177480C185592680 @default.
- W3085177480 hasConceptScore W3085177480C18903297 @default.
- W3085177480 hasConceptScore W3085177480C201995342 @default.
- W3085177480 hasConceptScore W3085177480C204321447 @default.
- W3085177480 hasConceptScore W3085177480C205649164 @default.
- W3085177480 hasConceptScore W3085177480C26517878 @default.
- W3085177480 hasConceptScore W3085177480C2778344882 @default.
- W3085177480 hasConceptScore W3085177480C2780451532 @default.
- W3085177480 hasConceptScore W3085177480C2780936489 @default.
- W3085177480 hasConceptScore W3085177480C38652104 @default.
- W3085177480 hasConceptScore W3085177480C41008148 @default.
- W3085177480 hasConceptScore W3085177480C47737302 @default.
- W3085177480 hasConceptScore W3085177480C86803240 @default.
- W3085177480 hasLocation W30851774801 @default.
- W3085177480 hasOpenAccess W3085177480 @default.
- W3085177480 hasPrimaryLocation W30851774801 @default.
- W3085177480 hasRelatedWork W2946659172 @default.
- W3085177480 hasRelatedWork W2963310665 @default.
- W3085177480 hasRelatedWork W2963341956 @default.
- W3085177480 hasRelatedWork W2963403868 @default.
- W3085177480 hasRelatedWork W2963748441 @default.
- W3085177480 hasRelatedWork W2964303116 @default.
- W3085177480 hasRelatedWork W2965373594 @default.
- W3085177480 hasRelatedWork W2970597249 @default.