{"id":"https://openalex.org/W7161154334","doi":"https://doi.org/10.48550/arxiv.2605.13370","title":"Phasor Memory Networks: Stable Backpropagation Through Time for Scalable Explicit Memory","display_name":"Phasor Memory Networks: Stable Backpropagation Through Time for Scalable Explicit Memory","publication_year":2026,"publication_date":"2026-05-13","ids":{"openalex":"https://openalex.org/W7161154334","doi":"https://doi.org/10.48550/arxiv.2605.13370"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.13370","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13370","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.13370","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5020854888","display_name":"Sungwoo Goo","orcid":"https://orcid.org/0000-0003-2809-5921"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Goo, Sungwoo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026070944","display_name":"Hwi\u2010yeol Yun","orcid":"https://orcid.org/0000-0001-8793-2449"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yun, Hwi-yeol","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136175077","display_name":"Sangkeun Jung","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jung, Sangkeun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.21150000393390656,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.21150000393390656,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12611","display_name":"Neural Networks and Reservoir Computing","score":0.13300000131130219,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.1315000057220459,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.5230000019073486},{"id":"https://openalex.org/keywords/phasor","display_name":"Phasor","score":0.5121999979019165},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.47540000081062317},{"id":"https://openalex.org/keywords/memristor","display_name":"Memristor","score":0.4740000069141388},{"id":"https://openalex.org/keywords/sequence-learning","display_name":"Sequence learning","score":0.4471000134944916},{"id":"https://openalex.org/keywords/memory-model","display_name":"Memory model","score":0.40880000591278076},{"id":"https://openalex.org/keywords/dynamic-random-access-memory","display_name":"Dynamic random-access memory","score":0.387800008058548},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.382999986410141}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7246999740600586},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.5230000019073486},{"id":"https://openalex.org/C176605952","wikidata":"https://www.wikidata.org/wiki/Q827674","display_name":"Phasor","level":4,"score":0.5121999979019165},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.47540000081062317},{"id":"https://openalex.org/C150072547","wikidata":"https://www.wikidata.org/wiki/Q212923","display_name":"Memristor","level":2,"score":0.4740000069141388},{"id":"https://openalex.org/C40506919","wikidata":"https://www.wikidata.org/wiki/Q7452469","display_name":"Sequence learning","level":2,"score":0.4471000134944916},{"id":"https://openalex.org/C12186640","wikidata":"https://www.wikidata.org/wiki/Q6815743","display_name":"Memory model","level":3,"score":0.40880000591278076},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.39649999141693115},{"id":"https://openalex.org/C118702147","wikidata":"https://www.wikidata.org/wiki/Q189396","display_name":"Dynamic random-access memory","level":3,"score":0.387800008058548},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.382999986410141},{"id":"https://openalex.org/C155032097","wikidata":"https://www.wikidata.org/wiki/Q798503","display_name":"Backpropagation","level":3,"score":0.3686000108718872},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.36649999022483826},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.3564999997615814},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.326200008392334},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32030001282691956},{"id":"https://openalex.org/C102392041","wikidata":"https://www.wikidata.org/wiki/Q592860","display_name":"Sliding window protocol","level":3,"score":0.31040000915527344},{"id":"https://openalex.org/C9870796","wikidata":"https://www.wikidata.org/wiki/Q490481","display_name":"Turing","level":2,"score":0.30630001425743103},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3046000003814697},{"id":"https://openalex.org/C179226034","wikidata":"https://www.wikidata.org/wiki/Q18599","display_name":"Short-term memory","level":4,"score":0.30239999294281006},{"id":"https://openalex.org/C171675096","wikidata":"https://www.wikidata.org/wiki/Q1143380","display_name":"Extended memory","level":4,"score":0.29910001158714294},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2903999984264374},{"id":"https://openalex.org/C20528329","wikidata":"https://www.wikidata.org/wiki/Q513932","display_name":"Linearizability","level":3,"score":0.257099986076355}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.13370","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13370","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.13370","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.13370","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.5139305591583252,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"For":[0],"over":[1],"a":[2,43,67,71,84,113,149,165,188,197],"decade,":[3],"explicit":[4,185],"memory":[5,49,110,124,186],"architectures":[6],"like":[7],"the":[8,97,105,109,139,160,181],"Neural":[9],"Turing":[10],"Machine":[11],"have":[12],"remained":[13],"theoretically":[14,198],"appealing":[15],"yet":[16],"practically":[17],"intractable":[18],"for":[19,99,201],"language":[20],"modeling":[21],"due":[22],"to":[23,80,127],"catastrophic":[24],"gradient":[25,90,177],"instability":[26],"during":[27],"Backpropagation":[28],"Through":[29],"Time.":[30],"In":[31],"this":[32,36],"work,":[33],"we":[34,65],"break":[35],"stalemate":[37],"with":[38],"\\textit{Phasor":[39],"Memory":[40],"Network}":[41],"(PMNet),":[42],"novel":[44],"architecture":[45],"that":[46,136,168,180],"structurally":[47],"resolves":[48],"volatility":[50],"through":[51,112],"\\textit{Unitary":[52],"Phasor":[53],"Dynamics}":[54],"and":[55,92,176],"\\textit{Hierarchical":[56],"Learnable":[57],"Anchors}.":[58],"Rather":[59],"than":[60],"relying":[61],"on":[62,83,155],"brute-force":[63],"scaling,":[64],"present":[66],"mechanistic":[68],"proof-of-concept":[69],"in":[70],"controlled":[72],"byte-level":[73],"setting.":[74],"By":[75],"constraining":[76],"recurrent":[77],"state":[78],"updates":[79],"phase":[81],"rotations":[82],"complex":[85],"unit":[86],"circle,":[87],"PMNet":[88,118,158,193],"preserves":[89],"norms":[91],"inherently":[93],"prevents":[94],"divergence":[95],"without":[96],"need":[98],"specialized":[100],"initialization.":[101],"We":[102],"empirically":[103],"demonstrate":[104],"active":[106],"actuation":[107],"of":[108,164,184],"module":[111],"synthetic":[114],"Copy-Paste":[115],"task,":[116],"where":[117],"utilizes":[119],"an":[120],"expansive":[121],"\\textit{85-slot":[122],"hierarchical":[123],"tree}":[125],"($=\\sum^{4}_{h=1}4^{h-1}$)":[126],"achieve":[128],"near":[129],"100\\%":[130],"exact":[131],"retrieval":[132],"across":[133],"temporal":[134],"distances":[135],"completely":[137],"exceed":[138],"local":[140],"sliding":[141],"window":[142],"attention's":[143],"receptive":[144],"field.":[145],"Furthermore,":[146],"despite":[147],"being":[148],"compact":[150],"119M":[151],"parameter":[152],"model":[153,167],"trained":[154],"18.8B":[156],"tokens,":[157],"matches":[159],"zero-shot":[161],"long-context":[162],"robustness":[163],"Mamba":[166],"is":[169],"three":[170],"times":[171],"larger.":[172],"Our":[173],"ablation":[174],"studies":[175],"analyses":[178],"confirm":[179],"historical":[182],"failure":[183],"was":[187],"structural":[189],"alignment":[190],"problem,":[191],"which":[192],"effectively":[194],"overcomes,":[195],"providing":[196],"grounded":[199],"foundation":[200],"scalable":[202],"sequence":[203],"modeling.":[204]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-15T00:00:00"}