Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310458839> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W4310458839 endingPage "7750" @default.
- W4310458839 startingPage "7739" @default.
- W4310458839 abstract "Recurrent neural networks are a widely used class of neural architectures. They have, however, two shortcomings. First, they are often treated as black-box models and as such it is difficult to understand what exactly they learn as well as how they arrive at a particular prediction. Second, they tend to work poorly on sequences requiring long-term memorization, despite having this capacity in principle. We aim to address both shortcomings with a class of recurrent networks that use a stochastic state transition mechanism between cell applications. This mechanism, which we term state-regularization, makes RNNs transition between a finite set of learnable states. We evaluate state-regularized RNNs on (1) regular languages for the purpose of automata extraction; (2) non-regular languages such as balanced parentheses and palindromes where external memory is required; and (3) real-word sequence learning tasks for sentiment analysis, visual object recognition and text categorisation. We show that state-regularization (a) simplifies the extraction of finite state automata that display an RNN's state transition dynamic; (b) forces RNNs to operate more like automata with external memory and less like finite state machines, which potentiality leads to a more structural memory; (c) leads to better interpretability and explainability of RNNs by leveraging the probabilistic finite state transition mechanism over time steps." @default.
- W4310458839 created "2022-12-10" @default.
- W4310458839 creator A5031719069 @default.
- W4310458839 creator A5061899878 @default.
- W4310458839 creator A5070442264 @default.
- W4310458839 date "2023-06-01" @default.
- W4310458839 modified "2023-09-24" @default.
- W4310458839 title "State-Regularized Recurrent Neural Networks to Extract Automata and Explain Predictions" @default.
- W4310458839 cites W1990813072 @default.
- W4310458839 cites W2005588250 @default.
- W4310458839 cites W2050778826 @default.
- W4310458839 cites W2064675550 @default.
- W4310458839 cites W2072160811 @default.
- W4310458839 cites W2098995689 @default.
- W4310458839 cites W2100649405 @default.
- W4310458839 cites W2112796928 @default.
- W4310458839 cites W2121029939 @default.
- W4310458839 cites W2134568112 @default.
- W4310458839 cites W2154039517 @default.
- W4310458839 cites W2530887700 @default.
- W4310458839 cites W2657631929 @default.
- W4310458839 cites W2764024122 @default.
- W4310458839 cites W2771330107 @default.
- W4310458839 cites W2963059228 @default.
- W4310458839 cites W2963363070 @default.
- W4310458839 cites W2963423043 @default.
- W4310458839 cites W2963921497 @default.
- W4310458839 cites W2964159778 @default.
- W4310458839 cites W4225680488 @default.
- W4310458839 cites W4244998381 @default.
- W4310458839 cites W4254816979 @default.
- W4310458839 doi "https://doi.org/10.1109/tpami.2022.3225334" @default.
- W4310458839 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/36445990" @default.
- W4310458839 hasPublicationYear "2023" @default.
- W4310458839 type Work @default.
- W4310458839 citedByCount "0" @default.
- W4310458839 crossrefType "journal-article" @default.
- W4310458839 hasAuthorship W4310458839A5031719069 @default.
- W4310458839 hasAuthorship W4310458839A5061899878 @default.
- W4310458839 hasAuthorship W4310458839A5070442264 @default.
- W4310458839 hasBestOaLocation W43104588392 @default.
- W4310458839 hasConcept C112505250 @default.
- W4310458839 hasConcept C11413529 @default.
- W4310458839 hasConcept C116248031 @default.
- W4310458839 hasConcept C147168706 @default.
- W4310458839 hasConcept C154945302 @default.
- W4310458839 hasConcept C167822520 @default.
- W4310458839 hasConcept C174327141 @default.
- W4310458839 hasConcept C2776135515 @default.
- W4310458839 hasConcept C2781067378 @default.
- W4310458839 hasConcept C41008148 @default.
- W4310458839 hasConcept C49937458 @default.
- W4310458839 hasConcept C50644808 @default.
- W4310458839 hasConcept C80444323 @default.
- W4310458839 hasConceptScore W4310458839C112505250 @default.
- W4310458839 hasConceptScore W4310458839C11413529 @default.
- W4310458839 hasConceptScore W4310458839C116248031 @default.
- W4310458839 hasConceptScore W4310458839C147168706 @default.
- W4310458839 hasConceptScore W4310458839C154945302 @default.
- W4310458839 hasConceptScore W4310458839C167822520 @default.
- W4310458839 hasConceptScore W4310458839C174327141 @default.
- W4310458839 hasConceptScore W4310458839C2776135515 @default.
- W4310458839 hasConceptScore W4310458839C2781067378 @default.
- W4310458839 hasConceptScore W4310458839C41008148 @default.
- W4310458839 hasConceptScore W4310458839C49937458 @default.
- W4310458839 hasConceptScore W4310458839C50644808 @default.
- W4310458839 hasConceptScore W4310458839C80444323 @default.
- W4310458839 hasIssue "6" @default.
- W4310458839 hasLocation W43104588391 @default.
- W4310458839 hasLocation W43104588392 @default.
- W4310458839 hasLocation W43104588393 @default.
- W4310458839 hasOpenAccess W4310458839 @default.
- W4310458839 hasPrimaryLocation W43104588391 @default.
- W4310458839 hasRelatedWork W1490980615 @default.
- W4310458839 hasRelatedWork W1979390054 @default.
- W4310458839 hasRelatedWork W1988501450 @default.
- W4310458839 hasRelatedWork W2044753318 @default.
- W4310458839 hasRelatedWork W2050696740 @default.
- W4310458839 hasRelatedWork W2076986528 @default.
- W4310458839 hasRelatedWork W2101451117 @default.
- W4310458839 hasRelatedWork W2159449605 @default.
- W4310458839 hasRelatedWork W2482365960 @default.
- W4310458839 hasRelatedWork W4310458839 @default.
- W4310458839 hasVolume "45" @default.
- W4310458839 isParatext "false" @default.
- W4310458839 isRetracted "false" @default.
- W4310458839 workType "article" @default.