Neuro-inspired cascaded memory coordination enables continual generalization

Neuro-inspired cascaded memory coordination enables continual generalization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Neuro-inspired cascaded memory coordination enables continual generalization Xingxing Zhang, Zijian Gao, Liyuan Wang, Bo Ding, Xinjun Mao, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9214055/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Biological systems continually incorporate new experiences while preserving prior knowledge through coordinated interactions across distributed memory systems. Artificial neural networks, by contrast, typically address continual learning by reducing interference, often through replay buffers or parameter isolation. Although effective for limiting forgetting, these strategies do not explicitly support the integration of new information into existing knowledge. Here we introduce the Cascaded Memory Network (CaMNet), a neuro-inspired continual learning framework that organizes memory into a progressive cascade of modules and uses compact cues to trigger targeted reinstatement of prior representations. Instead of retraining on stored samples, CaMNet stabilizes task-relevant knowledge by coordinating internal memory states and enforcing consistency across the cascade. Across continual vision–language and reinforcement learning benchmarks, CaMNet improves both retention and forward transfer relative to strong continual learning baselines. CaMNet also enables cue-driven recovery of latent memory traces after substantial performance degradation, indicating that previously acquired knowledge can be functionally re-engaged without revisiting prior inputs. In dynamic physical environments, the framework allows a humanoid robot to rapidly adapt to perceptual shifts and bridge the sim-to-real gap while preserving pre-trained capabilities. These results identify cascaded memory coordination as an effective computational principle for continual generalization and replay-free continual learning across domains. Physical sciences/Mathematics and computing/Computer science Physical sciences/Engineering Biological sciences/Neuroscience Continual Learning Continual Generalization Neuro-inspired AI Cascaded Memory Coordination Embodied AI Full Text Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryFileNeuroinspiredcascadedmemorycoordinationenablescontinualgeneralization.pdf Supplementary File for Neuro-inspired cascaded memory coordination enables continual generalization codeforNeuroinspiredcascadedmemorycoordinationenablescontinualgeneralization.zip Code and Dataset for Neuro-inspired cascaded memory coordination enables continual generalization Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9214055","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":627998164,"identity":"1c04fee0-da97-42dd-851a-eabe8c5cfd9c","order_by":0,"name":"Xingxing Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYHACNhDBw8/efvABiMFHrBYZyZ4zyQYgLWzEarExuJFgJgHn4gMGN9KfPfi5o5YHqCWt8muOnQwbA/PDRzfwaJHsOZBu2HvmOI/kmYfHbstuSwY6jM3YOAePFn72hmMSvG3HePiOJ6TdltzGDNTCwyaNTwsbM2Ob5F+gFoYDCWbFktvqCWvhZ29mk+Ztq+EROJFgxvhx22HCWiR7jrFJy7Yd4AEFsjTjtuM8bMwE/AIKMcm3bXX2oKj8+HNbNZDR/PAxPi1QcBhMMvOAScLKQaAOTDL+IE71KBgFo2AUjDAAAMOJRgtI8jFaAAAAAElFTkSuQmCC","orcid":"","institution":"Tsinghua University","correspondingAuthor":true,"prefix":"","firstName":"Xingxing","middleName":"","lastName":"Zhang","suffix":""},{"id":627998165,"identity":"3ed29854-a6a5-4077-94ba-2b52164a15c6","order_by":1,"name":"Zijian Gao","email":"","orcid":"","institution":"College of Computer Science and Technology, National University of Defense University","correspondingAuthor":false,"prefix":"","firstName":"Zijian","middleName":"","lastName":"Gao","suffix":""},{"id":627998166,"identity":"f999a245-64c1-4cde-b006-3ce23081caa9","order_by":2,"name":"Liyuan Wang","email":"","orcid":"https://orcid.org/0000-0002-3869-8155","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Liyuan","middleName":"","lastName":"Wang","suffix":""},{"id":627998167,"identity":"4e92a9fa-d372-4c31-9ec9-6c36a7867cf8","order_by":3,"name":"Bo Ding","email":"","orcid":"","institution":"College of Computer Science and Technology, National University of Defense University","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Ding","suffix":""},{"id":627998168,"identity":"2974c10c-eee9-4d8b-b144-53a795ddb663","order_by":4,"name":"Xinjun Mao","email":"","orcid":"","institution":"College of Computer Science and Technology, National University of Defense University","correspondingAuthor":false,"prefix":"","firstName":"Xinjun","middleName":"","lastName":"Mao","suffix":""},{"id":627998169,"identity":"89998148-4fa1-44d4-af15-fc8de95d6f52","order_by":5,"name":"Huaimin Wang","email":"","orcid":"","institution":"College of Computer Science and Technology, National University of Defense University","correspondingAuthor":false,"prefix":"","firstName":"Huaimin","middleName":"","lastName":"Wang","suffix":""},{"id":627998170,"identity":"83d6aecd-cf36-405c-915f-9980a850037e","order_by":6,"name":"Kele Xu","email":"","orcid":"","institution":"College of Computer Science and Technology, National University of Defense University","correspondingAuthor":false,"prefix":"","firstName":"Kele","middleName":"","lastName":"Xu","suffix":""}],"badges":[],"createdAt":"2026-03-24 15:43:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9214055/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9214055/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108493902,"identity":"00e40d71-d631-4e51-bbef-4bb843870cbd","added_by":"auto","created_at":"2026-05-05 10:02:07","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":20759857,"visible":true,"origin":"","legend":"Article File","description":"","filename":"Neuroinspiredcascadedmemorycoordinationenablescontinualgeneralization.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9214055/v1_covered_bfb425ca-f399-41be-bbb2-43f8618e09cb.pdf"},{"id":108446515,"identity":"a49bab84-9767-4e39-98bc-1daf8452969b","added_by":"auto","created_at":"2026-05-04 17:59:34","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":24922687,"visible":true,"origin":"","legend":"Supplementary File for Neuro-inspired cascaded memory coordination enables continual generalization","description":"","filename":"SupplementaryFileNeuroinspiredcascadedmemorycoordinationenablescontinualgeneralization.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9214055/v1/7971422139a458a68c411b66.pdf"},{"id":108446516,"identity":"8c57453a-1971-45e5-b565-135a5da516fd","added_by":"auto","created_at":"2026-05-04 17:59:34","extension":"zip","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":37752598,"visible":true,"origin":"","legend":"Code and Dataset for Neuro-inspired cascaded memory coordination enables continual generalization","description":"","filename":"codeforNeuroinspiredcascadedmemorycoordinationenablescontinualgeneralization.zip","url":"https://assets-eu.researchsquare.com/files/rs-9214055/v1/82e979d4d7acf1968d600a31.zip"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Neuro-inspired cascaded memory coordination enables continual generalization","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Continual Learning, Continual Generalization, Neuro-inspired AI, Cascaded Memory Coordination, Embodied AI","lastPublishedDoi":"10.21203/rs.3.rs-9214055/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9214055/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Biological systems continually incorporate new experiences while preserving prior knowledge through coordinated interactions across distributed memory systems. Artificial neural networks, by contrast, typically address continual learning by reducing interference, often through replay buffers or parameter isolation. Although effective for limiting forgetting, these strategies do not explicitly support the integration of new information into existing knowledge. Here we introduce the Cascaded Memory Network (CaMNet), a neuro-inspired continual learning framework that organizes memory into a progressive cascade of modules and uses compact cues to trigger targeted reinstatement of prior representations. Instead of retraining on stored samples, CaMNet stabilizes task-relevant knowledge by coordinating internal memory states and enforcing consistency across the cascade. Across continual vision–language and reinforcement learning benchmarks, CaMNet improves both retention and forward transfer relative to strong continual learning baselines. CaMNet also enables cue-driven recovery of latent memory traces after substantial performance degradation, indicating that previously acquired knowledge can be functionally re-engaged without revisiting prior inputs. In dynamic physical environments, the framework allows a humanoid robot to rapidly adapt to perceptual shifts and bridge the sim-to-real gap while preserving pre-trained capabilities. These results identify cascaded memory coordination as an effective computational principle for continual generalization and replay-free continual learning across domains.","manuscriptTitle":"Neuro-inspired cascaded memory coordination enables continual generalization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 17:59:29","doi":"10.21203/rs.3.rs-9214055/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"b2cd6074-ea74-4992-9d3c-166f93d39b6a","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":66843945,"name":"Physical sciences/Mathematics and computing/Computer science"},{"id":66843946,"name":"Physical sciences/Engineering"},{"id":66843947,"name":"Biological sciences/Neuroscience"}],"tags":[],"updatedAt":"2026-05-04T17:59:29+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 17:59:29","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9214055","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9214055","identity":"rs-9214055","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}