Matches in SemOpenAlex for { <https://semopenalex.org/work/W2742448943> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W2742448943 abstract "Recurrent neural networks (RNNs) serve as a fundamental building block for many sequence tasks across natural language processing. Recent research has focused on recurrent dropout techniques or custom RNN cells in order to improve performance. Both of these can require substantial modifications to the machine learning model or to the underlying RNN configurations. We revisit traditional regularization techniques, specifically L2 regularization on RNN activations and slowness regularization over successive hidden states, to improve the performance of RNNs on the task of language modeling. Both of these techniques require minimal modification to existing RNN architectures and result in performance improvements comparable or superior to more complicated regularization techniques or custom cell architectures. These regularization techniques can be used without any modification on optimized LSTM implementations such as the NVIDIA cuDNN LSTM." @default.
- W2742448943 created "2017-08-17" @default.
- W2742448943 creator A5004447350 @default.
- W2742448943 creator A5008147177 @default.
- W2742448943 creator A5059955534 @default.
- W2742448943 date "2017-08-03" @default.
- W2742448943 modified "2023-09-27" @default.
- W2742448943 title "Revisiting Activation Regularization for Language RNNs" @default.
- W2742448943 cites W1591801644 @default.
- W2742448943 cites W179875071 @default.
- W2742448943 cites W1800356822 @default.
- W2742448943 cites W2096388912 @default.
- W2742448943 cites W2124807415 @default.
- W2742448943 cites W2409027918 @default.
- W2742448943 cites W2473934411 @default.
- W2742448943 cites W2514713644 @default.
- W2742448943 cites W2525332836 @default.
- W2742448943 cites W2549416390 @default.
- W2742448943 cites W2553303224 @default.
- W2742448943 cites W2565031282 @default.
- W2742448943 cites W2949117887 @default.
- W2742448943 cites W2950297649 @default.
- W2742448943 cites W2950635152 @default.
- W2742448943 cites W2952013107 @default.
- W2742448943 cites W2952436057 @default.
- W2742448943 cites W2952507584 @default.
- W2742448943 cites W2963042606 @default.
- W2742448943 cites W2963266340 @default.
- W2742448943 hasPublicationYear "2017" @default.
- W2742448943 type Work @default.
- W2742448943 sameAs 2742448943 @default.
- W2742448943 citedByCount "25" @default.
- W2742448943 countsByYear W27424489432017 @default.
- W2742448943 countsByYear W27424489432018 @default.
- W2742448943 countsByYear W27424489432019 @default.
- W2742448943 countsByYear W27424489432020 @default.
- W2742448943 countsByYear W27424489432021 @default.
- W2742448943 crossrefType "posted-content" @default.
- W2742448943 hasAuthorship W2742448943A5004447350 @default.
- W2742448943 hasAuthorship W2742448943A5008147177 @default.
- W2742448943 hasAuthorship W2742448943A5059955534 @default.
- W2742448943 hasConcept C119857082 @default.
- W2742448943 hasConcept C137293760 @default.
- W2742448943 hasConcept C147168706 @default.
- W2742448943 hasConcept C154945302 @default.
- W2742448943 hasConcept C2776135515 @default.
- W2742448943 hasConcept C41008148 @default.
- W2742448943 hasConcept C50644808 @default.
- W2742448943 hasConceptScore W2742448943C119857082 @default.
- W2742448943 hasConceptScore W2742448943C137293760 @default.
- W2742448943 hasConceptScore W2742448943C147168706 @default.
- W2742448943 hasConceptScore W2742448943C154945302 @default.
- W2742448943 hasConceptScore W2742448943C2776135515 @default.
- W2742448943 hasConceptScore W2742448943C41008148 @default.
- W2742448943 hasConceptScore W2742448943C50644808 @default.
- W2742448943 hasLocation W27424489431 @default.
- W2742448943 hasOpenAccess W2742448943 @default.
- W2742448943 hasPrimaryLocation W27424489431 @default.
- W2742448943 hasRelatedWork W1591801644 @default.
- W2742448943 hasRelatedWork W1632114991 @default.
- W2742448943 hasRelatedWork W179875071 @default.
- W2742448943 hasRelatedWork W1836465849 @default.
- W2742448943 hasRelatedWork W2064675550 @default.
- W2742448943 hasRelatedWork W2095705004 @default.
- W2742448943 hasRelatedWork W2132339004 @default.
- W2742448943 hasRelatedWork W2194775991 @default.
- W2742448943 hasRelatedWork W2257784575 @default.
- W2742448943 hasRelatedWork W2284692871 @default.
- W2742448943 hasRelatedWork W2743945814 @default.
- W2742448943 hasRelatedWork W2792376130 @default.
- W2742448943 hasRelatedWork W2962964385 @default.
- W2742448943 hasRelatedWork W2963266340 @default.
- W2742448943 hasRelatedWork W2963347649 @default.
- W2742448943 hasRelatedWork W2963403868 @default.
- W2742448943 hasRelatedWork W2963537482 @default.
- W2742448943 hasRelatedWork W2963838731 @default.
- W2742448943 hasRelatedWork W2964308564 @default.
- W2742448943 hasRelatedWork W4919037 @default.
- W2742448943 isParatext "false" @default.
- W2742448943 isRetracted "false" @default.
- W2742448943 magId "2742448943" @default.
- W2742448943 workType "article" @default.