{"pk":31763,"title":"Jumpnet: A Multiple-Memory Connectionist Architecture","subtitle":null,"abstract":"A jumpnet includes two memory storage\nsystems: a processing network that employs\nsuperimpositional storage and a control\nnetwork that recodes input patterns into\nminimally overlapping hidden patterns. By\ncreating temporary, input-specific changes in\nthe weights of the processing network, the\ncontrol network causes the processing network to\n\"jvimp\" to the region of its weight space that is\nmost appropriate for a particular input pattern.\nSimulation results demonstrate that jumpnets\nexhibit only moderate levels of interference\nwhile retaining the computational advantages\nof superimpositional memory.","language":"eng","license":{"name":"","short_name":"","text":null,"url":""},"keywords":[],"section":"Submitted Presentations","is_remote":true,"remote_url":"https://escholarship.org/uc/item/2s03636r","frozenauthors":[{"first_name":"Jay","middle_name":"G.","last_name":"Rueckl","name_suffix":"","institution":"Harvard University","department":""}],"date_submitted":null,"date_accepted":null,"date_published":"1993-01-01T18:00:00Z","render_galley":null,"galleys":[{"label":"PDF","type":"pdf","path":"https://journalpub.escholarship.org/cognitivesciencesociety/article/31763/galley/22831/download/"}]}