Matches in SemOpenAlex for { <https://semopenalex.org/work/W3116353560> ?p ?o ?g. }
- W3116353560 abstract "Many concepts have been proposed for meta learning with neural networks (NNs), e.g., NNs that learn to control fast weights, hyper networks, learned learning rules, and meta recurrent NNs. Our Variable Shared Meta Learning (VS-ML) unifies the above and demonstrates that simple weight-sharing and sparsity in an NN is sufficient to express powerful learning algorithms (LAs) in a reusable fashion. A simple implementation of VS-ML called VS-ML RNN allows for implementing the backpropagation LA solely by running an RNN in forward-mode. It can even meta-learn new LAs that improve upon backpropagation and generalize to datasets outside of the meta training distribution without explicit gradient calculation. Introspection reveals that our meta-learned LAs learn qualitatively different from gradient descent through fast association." @default.
- W3116353560 created "2021-01-05" @default.
- W3116353560 creator A5010855168 @default.
- W3116353560 creator A5044629673 @default.
- W3116353560 date "2020-12-29" @default.
- W3116353560 modified "2023-09-27" @default.
- W3116353560 title "Meta Learning Backpropagation And Improving It" @default.
- W3116353560 cites W1598377843 @default.
- W3116353560 cites W185620388 @default.
- W3116353560 cites W1891300059 @default.
- W3116353560 cites W2064675550 @default.
- W3116353560 cites W2089217417 @default.
- W3116353560 cites W2102738208 @default.
- W3116353560 cites W2117085697 @default.
- W3116353560 cites W2136848157 @default.
- W3116353560 cites W2137825550 @default.
- W3116353560 cites W2144782183 @default.
- W3116353560 cites W2157331557 @default.
- W3116353560 cites W2166160300 @default.
- W3116353560 cites W2409744450 @default.
- W3116353560 cites W2472819217 @default.
- W3116353560 cites W2535697732 @default.
- W3116353560 cites W2550182557 @default.
- W3116353560 cites W2578206533 @default.
- W3116353560 cites W2593649546 @default.
- W3116353560 cites W2596367596 @default.
- W3116353560 cites W2604763608 @default.
- W3116353560 cites W2612506596 @default.
- W3116353560 cites W2734358244 @default.
- W3116353560 cites W2750384547 @default.
- W3116353560 cites W2753160622 @default.
- W3116353560 cites W2787501667 @default.
- W3116353560 cites W2891768589 @default.
- W3116353560 cites W2902986194 @default.
- W3116353560 cites W2906841252 @default.
- W3116353560 cites W2919053476 @default.
- W3116353560 cites W2952339051 @default.
- W3116353560 cites W2963168530 @default.
- W3116353560 cites W2963393838 @default.
- W3116353560 cites W2963775850 @default.
- W3116353560 cites W2964227899 @default.
- W3116353560 cites W2970292059 @default.
- W3116353560 cites W2979869797 @default.
- W3116353560 cites W2991922505 @default.
- W3116353560 cites W2995196103 @default.
- W3116353560 cites W2995686239 @default.
- W3116353560 cites W3001411605 @default.
- W3116353560 cites W3006201994 @default.
- W3116353560 cites W3010411571 @default.
- W3116353560 cites W3039213266 @default.
- W3116353560 cites W3043013488 @default.
- W3116353560 cites W3088321919 @default.
- W3116353560 cites W3101800035 @default.
- W3116353560 cites W3115727409 @default.
- W3116353560 cites W3117590783 @default.
- W3116353560 cites W3123651166 @default.
- W3116353560 cites W3138886030 @default.
- W3116353560 cites W3154346432 @default.
- W3116353560 hasPublicationYear "2020" @default.
- W3116353560 type Work @default.
- W3116353560 sameAs 3116353560 @default.
- W3116353560 citedByCount "7" @default.
- W3116353560 countsByYear W31163535602021 @default.
- W3116353560 countsByYear W31163535602022 @default.
- W3116353560 crossrefType "posted-content" @default.
- W3116353560 hasAuthorship W3116353560A5010855168 @default.
- W3116353560 hasAuthorship W3116353560A5044629673 @default.
- W3116353560 hasConcept C111472728 @default.
- W3116353560 hasConcept C119857082 @default.
- W3116353560 hasConcept C138885662 @default.
- W3116353560 hasConcept C147168706 @default.
- W3116353560 hasConcept C153258448 @default.
- W3116353560 hasConcept C154945302 @default.
- W3116353560 hasConcept C155032097 @default.
- W3116353560 hasConcept C162324750 @default.
- W3116353560 hasConcept C177973122 @default.
- W3116353560 hasConcept C187736073 @default.
- W3116353560 hasConcept C2780451532 @default.
- W3116353560 hasConcept C2780586882 @default.
- W3116353560 hasConcept C2781002164 @default.
- W3116353560 hasConcept C41008148 @default.
- W3116353560 hasConcept C50644808 @default.
- W3116353560 hasConcept C98359873 @default.
- W3116353560 hasConceptScore W3116353560C111472728 @default.
- W3116353560 hasConceptScore W3116353560C119857082 @default.
- W3116353560 hasConceptScore W3116353560C138885662 @default.
- W3116353560 hasConceptScore W3116353560C147168706 @default.
- W3116353560 hasConceptScore W3116353560C153258448 @default.
- W3116353560 hasConceptScore W3116353560C154945302 @default.
- W3116353560 hasConceptScore W3116353560C155032097 @default.
- W3116353560 hasConceptScore W3116353560C162324750 @default.
- W3116353560 hasConceptScore W3116353560C177973122 @default.
- W3116353560 hasConceptScore W3116353560C187736073 @default.
- W3116353560 hasConceptScore W3116353560C2780451532 @default.
- W3116353560 hasConceptScore W3116353560C2780586882 @default.
- W3116353560 hasConceptScore W3116353560C2781002164 @default.
- W3116353560 hasConceptScore W3116353560C41008148 @default.
- W3116353560 hasConceptScore W3116353560C50644808 @default.
- W3116353560 hasConceptScore W3116353560C98359873 @default.
- W3116353560 hasLocation W31163535601 @default.