Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952569483> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2952569483 abstract "Abstract A longstanding challenge for computational neuroscience has been the development of biologically plausible learning rules for recurrent neural networks (RNNs) enabling the production and processing of time-dependent signals such as those that might drive movement or facilitate working memory. Classic gradient-based algorithms for training RNNs have been available for decades, but they are inconsistent with known biological features of the brain, such as causality and locality. In this work we derive an approximation to gradient-based learning that comports with these biologically motivated constraints. Specifically, the online learning rule for the synaptic weights involves only local information about the pre- and postsynaptic activities, in addition to a random feedback projection of the RNN output error. In addition to providing mathematical arguments for the effectiveness of the new learning rule, we show through simulations that it can be used to train an RNN to successfully perform a variety of tasks. Finally, to overcome the difficulty of training an RNN over a very large number of timesteps, we propose an augmented circuit architecture that allows the RNN to concatenate short-duration patterns into sequences of longer duration." @default.
- W2952569483 created "2019-06-27" @default.
- W2952569483 creator A5025046799 @default.
- W2952569483 date "2018-10-31" @default.
- W2952569483 modified "2023-09-27" @default.
- W2952569483 title "Local online learning in recurrent networks with random feedback" @default.
- W2952569483 cites W1547537150 @default.
- W2952569483 cites W2008008116 @default.
- W2952569483 cites W2016589492 @default.
- W2952569483 cites W2019256317 @default.
- W2952569483 cites W2041176801 @default.
- W2952569483 cites W2047125104 @default.
- W2952569483 cites W2057409808 @default.
- W2952569483 cites W2063046089 @default.
- W2952569483 cites W2078569728 @default.
- W2952569483 cites W2080247083 @default.
- W2952569483 cites W2094037121 @default.
- W2952569483 cites W2099231026 @default.
- W2952569483 cites W2102701721 @default.
- W2952569483 cites W2105377180 @default.
- W2952569483 cites W2118706537 @default.
- W2952569483 cites W2118830830 @default.
- W2952569483 cites W2148201341 @default.
- W2952569483 cites W2219858099 @default.
- W2952569483 cites W2466948651 @default.
- W2952569483 cites W2529004582 @default.
- W2952569483 cites W2552737632 @default.
- W2952569483 cites W2611608598 @default.
- W2952569483 cites W2622558733 @default.
- W2952569483 cites W2766736793 @default.
- W2952569483 cites W2785323409 @default.
- W2952569483 cites W2785323625 @default.
- W2952569483 cites W2794448029 @default.
- W2952569483 cites W2804636914 @default.
- W2952569483 cites W2949405272 @default.
- W2952569483 cites W2953293408 @default.
- W2952569483 doi "https://doi.org/10.1101/458570" @default.
- W2952569483 hasPublicationYear "2018" @default.
- W2952569483 type Work @default.
- W2952569483 sameAs 2952569483 @default.
- W2952569483 citedByCount "0" @default.
- W2952569483 crossrefType "posted-content" @default.
- W2952569483 hasAuthorship W2952569483A5025046799 @default.
- W2952569483 hasBestOaLocation W29525694831 @default.
- W2952569483 hasConcept C112758219 @default.
- W2952569483 hasConcept C11413529 @default.
- W2952569483 hasConcept C119857082 @default.
- W2952569483 hasConcept C121332964 @default.
- W2952569483 hasConcept C124952713 @default.
- W2952569483 hasConcept C138885662 @default.
- W2952569483 hasConcept C142362112 @default.
- W2952569483 hasConcept C147168706 @default.
- W2952569483 hasConcept C154945302 @default.
- W2952569483 hasConcept C2779127903 @default.
- W2952569483 hasConcept C2779808786 @default.
- W2952569483 hasConcept C41008148 @default.
- W2952569483 hasConcept C41895202 @default.
- W2952569483 hasConcept C50644808 @default.
- W2952569483 hasConcept C57493831 @default.
- W2952569483 hasConcept C62520636 @default.
- W2952569483 hasConcept C64357122 @default.
- W2952569483 hasConceptScore W2952569483C112758219 @default.
- W2952569483 hasConceptScore W2952569483C11413529 @default.
- W2952569483 hasConceptScore W2952569483C119857082 @default.
- W2952569483 hasConceptScore W2952569483C121332964 @default.
- W2952569483 hasConceptScore W2952569483C124952713 @default.
- W2952569483 hasConceptScore W2952569483C138885662 @default.
- W2952569483 hasConceptScore W2952569483C142362112 @default.
- W2952569483 hasConceptScore W2952569483C147168706 @default.
- W2952569483 hasConceptScore W2952569483C154945302 @default.
- W2952569483 hasConceptScore W2952569483C2779127903 @default.
- W2952569483 hasConceptScore W2952569483C2779808786 @default.
- W2952569483 hasConceptScore W2952569483C41008148 @default.
- W2952569483 hasConceptScore W2952569483C41895202 @default.
- W2952569483 hasConceptScore W2952569483C50644808 @default.
- W2952569483 hasConceptScore W2952569483C57493831 @default.
- W2952569483 hasConceptScore W2952569483C62520636 @default.
- W2952569483 hasConceptScore W2952569483C64357122 @default.
- W2952569483 hasLocation W29525694831 @default.
- W2952569483 hasLocation W29525694832 @default.
- W2952569483 hasOpenAccess W2952569483 @default.
- W2952569483 hasPrimaryLocation W29525694831 @default.
- W2952569483 hasRelatedWork W1523525818 @default.
- W2952569483 hasRelatedWork W2793022090 @default.
- W2952569483 hasRelatedWork W2899217980 @default.
- W2952569483 hasRelatedWork W2969748604 @default.
- W2952569483 hasRelatedWork W3192794374 @default.
- W2952569483 hasRelatedWork W4200442073 @default.
- W2952569483 hasRelatedWork W4205256556 @default.
- W2952569483 hasRelatedWork W4213142596 @default.
- W2952569483 hasRelatedWork W4255998495 @default.
- W2952569483 hasRelatedWork W4281386417 @default.
- W2952569483 isParatext "false" @default.
- W2952569483 isRetracted "false" @default.
- W2952569483 magId "2952569483" @default.
- W2952569483 workType "article" @default.