FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3128750371> ?p ?o ?g. }

• W3128750371 endingPage "3621" @default.
• W3128750371 startingPage "3612" @default.
• W3128750371 abstract "Attention has been shown highly effective for modeling sequences, capturing the more informative parts in learning a deep representation. However, recent studies show that the attention values do not always coincide with intuition in tasks, such as machine translation and sentiment classification. In this study, we consider using deep reinforcement learning to automatically optimize attention distribution during the minimization of end task training losses. With more sufficient environment states, iterative actions are taken to adjust attention weights so that more informative words receive more attention automatically. Results on different tasks and different attention networks demonstrate that our model is of great effectiveness in improving the end task performances, yielding more reasonable attention distribution. The more in-depth analysis further reveals that our retrofitting method can help to bring explainability for baseline attention." @default.
• W3128750371 created "2021-02-15" @default.
• W3128750371 creator A5002241280 @default.
• W3128750371 creator A5055938662 @default.
• W3128750371 creator A5058877618 @default.
• W3128750371 creator A5089500802 @default.
• W3128750371 date "2022-08-01" @default.
• W3128750371 modified "2023-10-18" @default.
• W3128750371 title "Optimizing Attention for Sequence Modeling via Reinforcement Learning" @default.
• W3128750371 cites W1832693441 @default.
• W3128750371 cites W1902237438 @default.
• W3128750371 cites W2024390895 @default.
• W3128750371 cites W2064675550 @default.
• W3128750371 cites W2114524997 @default.
• W3128750371 cites W2119717200 @default.
• W3128750371 cites W2163455955 @default.
• W3128750371 cites W2562607067 @default.
• W3128750371 cites W2563010554 @default.
• W3128750371 cites W2604205681 @default.
• W3128750371 cites W2757154661 @default.
• W3128750371 cites W2785128315 @default.
• W3128750371 cites W2805151692 @default.
• W3128750371 cites W2807741081 @default.
• W3128750371 cites W2808182015 @default.
• W3128750371 cites W2895944290 @default.
• W3128750371 cites W2950768109 @default.
• W3128750371 cites W2952028139 @default.
• W3128750371 cites W2952406142 @default.
• W3128750371 cites W2962729168 @default.
• W3128750371 cites W2963051485 @default.
• W3128750371 cites W2963467630 @default.
• W3128750371 cites W2963499882 @default.
• W3128750371 cites W2963696295 @default.
• W3128750371 cites W2963888230 @default.
• W3128750371 cites W2963963856 @default.
• W3128750371 cites W2964005754 @default.
• W3128750371 cites W2964164368 @default.
• W3128750371 cites W2964268978 @default.
• W3128750371 cites W2970120757 @default.
• W3128750371 cites W2970726176 @default.
• W3128750371 cites W2971351900 @default.
• W3128750371 cites W2977700998 @default.
• W3128750371 cites W2989149555 @default.
• W3128750371 cites W2997003477 @default.
• W3128750371 cites W2997087088 @default.
• W3128750371 cites W3034444624 @default.
• W3128750371 cites W3104213339 @default.
• W3128750371 cites W2952802110 @default.
• W3128750371 doi "https://doi.org/10.1109/tnnls.2021.3053633" @default.
• W3128750371 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/33566767" @default.
• W3128750371 hasPublicationYear "2022" @default.
• W3128750371 type Work @default.
• W3128750371 sameAs 3128750371 @default.
• W3128750371 citedByCount "7" @default.
• W3128750371 countsByYear W31287503712021 @default.
• W3128750371 countsByYear W31287503712022 @default.
• W3128750371 countsByYear W31287503712023 @default.
• W3128750371 crossrefType "journal-article" @default.
• W3128750371 hasAuthorship W3128750371A5002241280 @default.
• W3128750371 hasAuthorship W3128750371A5055938662 @default.
• W3128750371 hasAuthorship W3128750371A5058877618 @default.
• W3128750371 hasAuthorship W3128750371A5089500802 @default.
• W3128750371 hasConcept C119857082 @default.
• W3128750371 hasConcept C127413603 @default.
• W3128750371 hasConcept C132010649 @default.
• W3128750371 hasConcept C154945302 @default.
• W3128750371 hasConcept C15744967 @default.
• W3128750371 hasConcept C188147891 @default.
• W3128750371 hasConcept C201995342 @default.
• W3128750371 hasConcept C203005215 @default.
• W3128750371 hasConcept C2780451532 @default.
• W3128750371 hasConcept C41008148 @default.
• W3128750371 hasConcept C97541855 @default.
• W3128750371 hasConceptScore W3128750371C119857082 @default.
• W3128750371 hasConceptScore W3128750371C127413603 @default.
• W3128750371 hasConceptScore W3128750371C132010649 @default.
• W3128750371 hasConceptScore W3128750371C154945302 @default.
• W3128750371 hasConceptScore W3128750371C15744967 @default.
• W3128750371 hasConceptScore W3128750371C188147891 @default.
• W3128750371 hasConceptScore W3128750371C201995342 @default.
• W3128750371 hasConceptScore W3128750371C203005215 @default.
• W3128750371 hasConceptScore W3128750371C2780451532 @default.
• W3128750371 hasConceptScore W3128750371C41008148 @default.
• W3128750371 hasConceptScore W3128750371C97541855 @default.
• W3128750371 hasFunder F4320321001 @default.
• W3128750371 hasFunder F4320336751 @default.
• W3128750371 hasIssue "8" @default.
• W3128750371 hasLocation W31287503711 @default.
• W3128750371 hasLocation W31287503712 @default.
• W3128750371 hasOpenAccess W3128750371 @default.
• W3128750371 hasPrimaryLocation W31287503711 @default.
• W3128750371 hasRelatedWork W1963934847 @default.
• W3128750371 hasRelatedWork W1974878518 @default.
• W3128750371 hasRelatedWork W2923653485 @default.
• W3128750371 hasRelatedWork W2957776456 @default.
• W3128750371 hasRelatedWork W3022038857 @default.
• W3128750371 hasRelatedWork W3107474891 @default.
• W3128750371 hasRelatedWork W4206493799 @default.

• next