Brain-Guided Convolutional Neural Networks Reveal Task-Specific Representations in Scene Processing | 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 Brain-Guided Convolutional Neural Networks Reveal Task-Specific Representations in Scene Processing Bruce C. Hansen, Michelle R. Greene, Henry A.S. Lewinsohn, Audrey E. Kris, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5753306/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Apr, 2025 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Scene categorization is the dominant proxy for visual understanding, yet humans can perform a large number of visual tasks within any scene. Consequently, we know little about how different tasks change how the is scene processed, represented, and its features ultimately used. Here, we developed a novel brain-guided convolutional neural network (CNN) where each convolutional layer was separately guided by neural responses taken at different time points while observers performed two different tasks on the same set of images. We then reconstructed each layer’s activation maps via deconvolution to spatially assess how different features were used as a function of task. The brain-guided CNN made use of image features that human observers identified as being crucial to complete each task starting around 244 ms and persisted to 402 ms. Critically, because the same images were used across the two tasks, the CNN could only succeed if the neural data captured task-relevant differences. Our analyses of the activation maps across layers revealed that the brain’s spatiotemporal representation of local image features evolves systematically over time. This underscores how distinct image features emerge at different stages of processing, shaped by the observer’s goals and behavioral context. Biological sciences/Neuroscience Biological sciences/Neuroscience/Cognitive neuroscience Biological sciences/Neuroscience/Computational neuroscience Biological sciences/Neuroscience/Visual system Scene understanding electroencephalography (EEG) convolutional neural networks (CNN) brain-guided neural networks Full Text Additional Declarations No competing interests reported. Supplementary Files HansenGreeneetalSupplementalMaterial.docx Cite Share Download PDF Status: Published Journal Publication published 15 Apr, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 03 Feb, 2025 Reviews received at journal 31 Jan, 2025 Reviews received at journal 30 Jan, 2025 Reviewers agreed at journal 22 Jan, 2025 Reviewers agreed at journal 20 Jan, 2025 Reviewers invited by journal 20 Jan, 2025 Editor assigned by journal 20 Jan, 2025 Editor invited by journal 09 Jan, 2025 Submission checks completed at journal 09 Jan, 2025 First submitted to journal 02 Jan, 2025 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-5753306","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":400017662,"identity":"19c77e1b-7bc5-4974-b713-2d4d0e18ba54","order_by":0,"name":"Bruce C. Hansen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6ElEQVRIiWNgGAWjYBACAwYeCIOfgYHxAFyQKC2SDQwMB6CqidRicIBYLebsZw9+Lvhjk2d8/PCBg19q/sgxsDdvk8CnxbInL1l6ZltasdmZtITDMscMjBl4jpXh1WJwg8dAmrfhcOI2IOOwBJtBYoNEjhkhLca/ef78T9w8A6Tln0F9g/wbglrMpHnYDiRukOAxOPixzSCBQYIHvxbLnhwza9625MQZIL8w9hkbtvGkFVvg02LOfsb4Ns8fu8T+9sMHH/74JifPz3544w18WlAAMyiO2IhWDgKMP0hSPgpGwSgYBSMFAAC3+Uijhne51wAAAABJRU5ErkJggg==","orcid":"","institution":"Colgate University","correspondingAuthor":true,"prefix":"","firstName":"Bruce","middleName":"C.","lastName":"Hansen","suffix":""},{"id":400017663,"identity":"9d7fe5af-1885-49a8-9c33-db61a7ee5c08","order_by":1,"name":"Michelle R. Greene","email":"","orcid":"","institution":"Barnard College","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"R.","lastName":"Greene","suffix":""},{"id":400017664,"identity":"592fba9d-9495-4762-b024-6e6449a25149","order_by":2,"name":"Henry A.S. Lewinsohn","email":"","orcid":"","institution":"Colgate University","correspondingAuthor":false,"prefix":"","firstName":"Henry","middleName":"A.S.","lastName":"Lewinsohn","suffix":""},{"id":400017665,"identity":"300c46c1-1cb0-4e53-93bb-63eb382f178a","order_by":3,"name":"Audrey E. Kris","email":"","orcid":"","institution":"Colgate University","correspondingAuthor":false,"prefix":"","firstName":"Audrey","middleName":"E.","lastName":"Kris","suffix":""},{"id":400017667,"identity":"1ebd82e2-dff8-447e-b22c-40002c5665d9","order_by":4,"name":"Sophie Smyth","email":"","orcid":"","institution":"Colgate University","correspondingAuthor":false,"prefix":"","firstName":"Sophie","middleName":"","lastName":"Smyth","suffix":""},{"id":400017668,"identity":"c58fb2fc-2a65-4b58-9516-32ecd647b1c4","order_by":5,"name":"Binghui Tang","email":"","orcid":"","institution":"Colgate University","correspondingAuthor":false,"prefix":"","firstName":"Binghui","middleName":"","lastName":"Tang","suffix":""}],"badges":[],"createdAt":"2025-01-02 16:53:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5753306/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5753306/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-96307-w","type":"published","date":"2025-04-15T15:56:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81050799,"identity":"8306ab59-6d6c-45a8-b172-d0a9afb53efb","added_by":"auto","created_at":"2025-04-21 16:05:19","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1162991,"visible":true,"origin":"","legend":"","description":"","filename":"HansenGreeneetalManuscriptFile.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5753306/v1_covered_965b7bb5-9c06-476a-91f3-7624acc7cd9a.pdf"},{"id":73629040,"identity":"bd21837b-6c9a-4fad-8e07-52d0aad3e41e","added_by":"auto","created_at":"2025-01-13 06:07:25","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":576707,"visible":true,"origin":"","legend":"","description":"","filename":"HansenGreeneetalSupplementalMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-5753306/v1/79c1b948a296dd10ac451c5f.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Brain-Guided Convolutional Neural Networks Reveal Task-Specific Representations in Scene Processing","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"
[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Scene understanding, electroencephalography (EEG), convolutional neural networks (CNN), brain-guided neural networks","lastPublishedDoi":"10.21203/rs.3.rs-5753306/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5753306/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Scene categorization is the dominant proxy for visual understanding, yet humans can perform a large number of visual tasks within any scene. Consequently, we know little about how different tasks change how the is scene processed, represented, and its features ultimately used. Here, we developed a novel brain-guided convolutional neural network (CNN) where each convolutional layer was separately guided by neural responses taken at different time points while observers performed two different tasks on the same set of images. We then reconstructed each layer’s activation maps via deconvolution to spatially assess how different features were used as a function of task. The brain-guided CNN made use of image features that human observers identified as being crucial to complete each task starting around 244 ms and persisted to 402 ms. Critically, because the same images were used across the two tasks, the CNN could only succeed if the neural data captured task-relevant differences. Our analyses of the activation maps across layers revealed that the brain’s spatiotemporal representation of local image features evolves systematically over time. This underscores how distinct image features emerge at different stages of processing, shaped by the observer’s goals and behavioral context.","manuscriptTitle":"Brain-Guided Convolutional Neural Networks Reveal Task-Specific Representations in Scene Processing","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-13 06:07:20","doi":"10.21203/rs.3.rs-5753306/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-02-03T14:37:56+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-01-31T17:27:31+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-01-30T21:07:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"88181672222519061484013293004925320945","date":"2025-01-22T17:38:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"104245639241855336826504706660215205975","date":"2025-01-20T23:35:55+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-01-20T18:58:49+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-20T18:58:01+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-01-09T14:36:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-09T14:35:29+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-01-02T16:46:38+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"
[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e819ccae-5d00-47f2-8c74-d2682189fd27","owner":[],"postedDate":"January 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42630519,"name":"Biological sciences/Neuroscience"},{"id":42630520,"name":"Biological sciences/Neuroscience/Cognitive neuroscience"},{"id":42630521,"name":"Biological sciences/Neuroscience/Computational neuroscience"},{"id":42630522,"name":"Biological sciences/Neuroscience/Visual system"}],"tags":[],"updatedAt":"2025-04-21T15:59:50+00:00","versionOfRecord":{"articleIdentity":"rs-5753306","link":"https://doi.org/10.1038/s41598-025-96307-w","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-04-15 15:56:57","publishedOnDateReadable":"April 15th, 2025"},"versionCreatedAt":"2025-01-13 06:07:20","video":"","vorDoi":"10.1038/s41598-025-96307-w","vorDoiUrl":"https://doi.org/10.1038/s41598-025-96307-w","workflowStages":[]},"version":"v1","identity":"rs-5753306","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5753306","identity":"rs-5753306","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.