Mostrar el registro sencillo del ítem
dc.contributor.author
Jarne, Cecilia Gisele
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.contributor.author
Laje, Rodrigo
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.date.available
2023-12-19T11:39:47Z
dc.date.issued
2023-08
dc.identifier.citation
Jarne, Cecilia Gisele; Laje, Rodrigo; Exploring weight initialization, diversity of solutions, and degradation in recurrent neural networks trained for temporal and decision-making tasks; Springer; Journal of Computational Neuroscience; 51; 4; 8-2023; 407-431
dc.identifier.issn
0929-5313
dc.identifier.uri
http://hdl.handle.net/11336/220736
dc.description.abstract
Recurrent Neural Networks (RNNs) are frequently used to model aspects of brain function and structure. In this work, we trained small fully-connected RNNs to perform temporal and flow control tasks with time-varying stimuli. Our results show that different RNNs can solve the same task by converging to different underlying dynamics and also how the performance gracefully degrades as either network size is decreased, interval duration is increased, or connectivity damage is induced. For the considered tasks, we explored how robust the network obtained after training can be according to task parameterization. In the process, we developed a framework that can be useful to parameterize other tasks of interest in computational neuroscience. Our results are useful to quantify different aspects of the models, which are normally used as black boxes and need to be understood in order to model the biological response of cerebral cortex areas.
dc.format
application/pdf
dc.language.iso
eng
dc.publisher
Springer
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.rights
info:eu-repo/semantics/restrictedAccess
dc.rights.uri
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.subject
CONNECTIVITY
dc.subject
DEGRADATION
dc.subject
RECURRENT NEURAL NETWORKS
dc.subject.classification
Otras Ciencias Físicas
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.subject.classification
Ciencias Físicas
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.subject.classification
CIENCIAS NATURALES Y EXACTAS
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.title
Exploring weight initialization, diversity of solutions, and degradation in recurrent neural networks trained for temporal and decision-making tasks
dc.type
info:eu-repo/semantics/article
dc.type
info:ar-repo/semantics/artículo
dc.type
info:eu-repo/semantics/publishedVersion
dc.date.updated
2023-12-12T15:43:59Z
dc.journal.volume
51
dc.journal.number
4
dc.journal.pagination
407-431
dc.journal.pais
Alemania
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.description.fil
Fil: Jarne, Cecilia Gisele. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Quilmes. Departamento de Ciencia y Tecnología; Argentina
dc.description.fil
Fil: Laje, Rodrigo. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Quilmes. Departamento de Ciencia y Tecnología; Argentina
dc.journal.title
Journal of Computational Neuroscience
![Se ha confirmado la validez de este valor de autoridad por un usuario](/themes/CONICETDigital/images/authority_control/invisible.gif)
dc.relation.alternativeid
info:eu-repo/semantics/altIdentifier/url/https://link.springer.com/10.1007/s10827-023-00857-9
dc.relation.alternativeid
info:eu-repo/semantics/altIdentifier/doi/http://dx.doi.org/10.1007/s10827-023-00857-9
Archivos asociados