Evaluating the perception, understanding, and forgetting of Progressive Neural Networks: a quantitative and qualitative analysis

Evaluating the perception, understanding, and forgetting of Progressive Neural Networks: a quantitative and qualitative analysis | 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 Research Article Evaluating the perception, understanding, and forgetting of Progressive Neural Networks: a quantitative and qualitative analysis Lucía Güitta-López, Jaime Boal, Álvaro Jesús López-López This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6662623/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The use of virtual environments to collect the experience required by deep reinforcement learning models is accelerating the deployment of these algorithms in industrial environments. However, once the experience-gathering problem is solved, it is necessary to address how to efficiently transfer the knowledge from the virtual scenario to reality. This paper focuses on examining Progressive Neural Networks (PNNs) as a promising transfer learning technique. The analyses carried out range from studying the capabilities and limits of the layers in charge of learning the state representation from a pixel space, which could arguably be the convolutional blocks, to the forgetting agents suffer when learning a new task. Introducing controlled visual changes in the environment scene can lead to a performance degradation of up to 50%. These visual discrepancies strongly influence the agent learning time and its accuracy when using a PNN architecture. Regarding the PNN forgetting assessment, partial forgetting occurs in two of the three environments analyzed, those where the agent masters its new task. This could be due to a balance between the relevance of the new features learned and the ones inherited from the teacher agent. Artificial Intelligence and Machine Learning Robotics Deep Reinforcement Learning Progressive Neural Networks Sim-To-Real Sample efficiency Representation Learning Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted 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-6662623","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":456433166,"identity":"a3574cf9-e3ed-421a-a8c0-772fb0545dda","order_by":0,"name":"Lucía Güitta-López","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0003-2426-3748","institution":"Comillas Pontifical University","correspondingAuthor":true,"prefix":"","firstName":"Lucía","middleName":"","lastName":"Güitta-López","suffix":""},{"id":456433337,"identity":"27aa1753-5c46-4730-98ea-dfb11db60355","order_by":1,"name":"Jaime Boal","email":"","orcid":"https://orcid.org/0000-0002-7547-0942","institution":"Comillas Pontifical University","correspondingAuthor":false,"prefix":"","firstName":"Jaime","middleName":"","lastName":"Boal","suffix":""},{"id":456433338,"identity":"f125ef41-76c0-4386-95cd-212b41773660","order_by":2,"name":"Álvaro Jesús López-López","email":"","orcid":"https://orcid.org/0000-0001-9879-5603","institution":"Comillas Pontifical University","correspondingAuthor":false,"prefix":"","firstName":"Álvaro","middleName":"Jesús","lastName":"López-López","suffix":""}],"badges":[],"createdAt":"2025-05-14 09:23:37","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6662623/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6662623/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82782671,"identity":"0f23a870-2b2c-430b-af7a-176c63b52624","added_by":"auto","created_at":"2025-05-15 08:38:29","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":851994,"visible":true,"origin":"","legend":"","description":"","filename":"snarticle.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6662623/v1_covered_f440ec84-163d-4bfc-ae47-8ffeea78935f.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eEvaluating the perception, understanding, and forgetting of Progressive Neural Networks: a quantitative and qualitative analysis\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Comillas Pontifical University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Deep Reinforcement Learning, Progressive Neural Networks, Sim-To-Real, Sample efficiency, Representation Learning","lastPublishedDoi":"10.21203/rs.3.rs-6662623/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6662623/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe use of virtual environments to collect the experience required by deep reinforcement learning models is accelerating the deployment of these algorithms in industrial environments. However, once the experience-gathering problem is solved, it is necessary to address how to efficiently transfer the knowledge from the virtual scenario to reality. This paper focuses on examining Progressive Neural Networks (PNNs) as a promising transfer learning technique. The analyses carried out range from studying the capabilities and limits of the layers in charge of learning the state representation from a pixel space, which could arguably be the convolutional blocks, to the forgetting agents suffer when learning a new task. Introducing controlled visual changes in the environment scene can lead to a performance degradation of up to 50%. These visual discrepancies strongly influence the agent learning time and its accuracy when using a PNN architecture. Regarding the PNN forgetting assessment, partial forgetting occurs in two of the three environments analyzed, those where the agent masters its new task. This could be due to a balance between the relevance of the new features learned and the ones inherited from the teacher agent.\u003c/p\u003e","manuscriptTitle":"Evaluating the perception, understanding, and forgetting of Progressive Neural Networks: a quantitative and qualitative analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-15 08:30:21","doi":"10.21203/rs.3.rs-6662623/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d62f4ea0-f781-42e9-941e-446415024ea5","owner":[],"postedDate":"May 15th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48513912,"name":"Artificial Intelligence and Machine Learning"},{"id":48513913,"name":"Robotics"}],"tags":[],"updatedAt":"2025-05-15T08:30:21+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-15 08:30:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6662623","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6662623","identity":"rs-6662623","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}