Matches in SemOpenAlex for { <https://semopenalex.org/work/W2805024782> ?p ?o ?g. }
- W2805024782 abstract "Word embedding parameters often dominate overall model sizes in neural methods for natural language processing. We reduce deployed model sizes of text classifiers by learning a hard word clustering in an end-to-end manner. We use the Gumbel-Softmax distribution to maximize over the latent clustering while minimizing the task loss. We propose variations that selectively assign additional parameters to words, which further improves accuracy while still remaining parameter-efficient." @default.
- W2805024782 created "2018-06-01" @default.
- W2805024782 creator A5070933659 @default.
- W2805024782 creator A5081022650 @default.
- W2805024782 date "2018-01-01" @default.
- W2805024782 modified "2023-09-26" @default.
- W2805024782 title "Smaller Text Classifiers with Discriminative Cluster Embeddings" @default.
- W2805024782 cites W1832693441 @default.
- W2805024782 cites W2064675550 @default.
- W2805024782 cites W2113459411 @default.
- W2805024782 cites W2121227244 @default.
- W2805024782 cites W2132339004 @default.
- W2805024782 cites W2250539671 @default.
- W2805024782 cites W2547875792 @default.
- W2805024782 cites W2548228487 @default.
- W2805024782 cites W2563351168 @default.
- W2805024782 cites W2962790997 @default.
- W2805024782 cites W2962943936 @default.
- W2805024782 cites W2963012544 @default.
- W2805024782 cites W2963424553 @default.
- W2805024782 cites W2963674932 @default.
- W2805024782 cites W2964121744 @default.
- W2805024782 cites W2964299589 @default.
- W2805024782 doi "https://doi.org/10.18653/v1/n18-2116" @default.
- W2805024782 hasPublicationYear "2018" @default.
- W2805024782 type Work @default.
- W2805024782 sameAs 2805024782 @default.
- W2805024782 citedByCount "5" @default.
- W2805024782 countsByYear W28050247822019 @default.
- W2805024782 countsByYear W28050247822020 @default.
- W2805024782 crossrefType "proceedings-article" @default.
- W2805024782 hasAuthorship W2805024782A5070933659 @default.
- W2805024782 hasAuthorship W2805024782A5081022650 @default.
- W2805024782 hasBestOaLocation W28050247821 @default.
- W2805024782 hasConcept C105795698 @default.
- W2805024782 hasConcept C119857082 @default.
- W2805024782 hasConcept C137293760 @default.
- W2805024782 hasConcept C137610916 @default.
- W2805024782 hasConcept C147581598 @default.
- W2805024782 hasConcept C153180895 @default.
- W2805024782 hasConcept C154945302 @default.
- W2805024782 hasConcept C162324750 @default.
- W2805024782 hasConcept C164866538 @default.
- W2805024782 hasConcept C187736073 @default.
- W2805024782 hasConcept C188441871 @default.
- W2805024782 hasConcept C199360897 @default.
- W2805024782 hasConcept C204321447 @default.
- W2805024782 hasConcept C2524010 @default.
- W2805024782 hasConcept C2777462759 @default.
- W2805024782 hasConcept C2780451532 @default.
- W2805024782 hasConcept C33923547 @default.
- W2805024782 hasConcept C41008148 @default.
- W2805024782 hasConcept C41608201 @default.
- W2805024782 hasConcept C50644808 @default.
- W2805024782 hasConcept C73555534 @default.
- W2805024782 hasConcept C90805587 @default.
- W2805024782 hasConcept C97931131 @default.
- W2805024782 hasConceptScore W2805024782C105795698 @default.
- W2805024782 hasConceptScore W2805024782C119857082 @default.
- W2805024782 hasConceptScore W2805024782C137293760 @default.
- W2805024782 hasConceptScore W2805024782C137610916 @default.
- W2805024782 hasConceptScore W2805024782C147581598 @default.
- W2805024782 hasConceptScore W2805024782C153180895 @default.
- W2805024782 hasConceptScore W2805024782C154945302 @default.
- W2805024782 hasConceptScore W2805024782C162324750 @default.
- W2805024782 hasConceptScore W2805024782C164866538 @default.
- W2805024782 hasConceptScore W2805024782C187736073 @default.
- W2805024782 hasConceptScore W2805024782C188441871 @default.
- W2805024782 hasConceptScore W2805024782C199360897 @default.
- W2805024782 hasConceptScore W2805024782C204321447 @default.
- W2805024782 hasConceptScore W2805024782C2524010 @default.
- W2805024782 hasConceptScore W2805024782C2777462759 @default.
- W2805024782 hasConceptScore W2805024782C2780451532 @default.
- W2805024782 hasConceptScore W2805024782C33923547 @default.
- W2805024782 hasConceptScore W2805024782C41008148 @default.
- W2805024782 hasConceptScore W2805024782C41608201 @default.
- W2805024782 hasConceptScore W2805024782C50644808 @default.
- W2805024782 hasConceptScore W2805024782C73555534 @default.
- W2805024782 hasConceptScore W2805024782C90805587 @default.
- W2805024782 hasConceptScore W2805024782C97931131 @default.
- W2805024782 hasLocation W28050247821 @default.
- W2805024782 hasLocation W28050247822 @default.
- W2805024782 hasOpenAccess W2805024782 @default.
- W2805024782 hasPrimaryLocation W28050247821 @default.
- W2805024782 hasRelatedWork W134561036 @default.
- W2805024782 hasRelatedWork W134865995 @default.
- W2805024782 hasRelatedWork W1994859265 @default.
- W2805024782 hasRelatedWork W2150099067 @default.
- W2805024782 hasRelatedWork W2159859968 @default.
- W2805024782 hasRelatedWork W2228274226 @default.
- W2805024782 hasRelatedWork W2250539671 @default.
- W2805024782 hasRelatedWork W2512569467 @default.
- W2805024782 hasRelatedWork W2516723902 @default.
- W2805024782 hasRelatedWork W2759863041 @default.
- W2805024782 hasRelatedWork W2942985259 @default.
- W2805024782 hasRelatedWork W2958981444 @default.
- W2805024782 hasRelatedWork W2963477629 @default.
- W2805024782 hasRelatedWork W2970666910 @default.
- W2805024782 hasRelatedWork W2984693480 @default.
- W2805024782 hasRelatedWork W3042751648 @default.