Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence

Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence | 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 Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence DIDIER Grimaldi, Carlos Carrasco-Farré This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9161370/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 As LLMs increasingly mediate decisions, recommendations, and persuasive communication, a central question is whether their mind-attribution structure in output space is human-like or only behaviorally similar. This creates a behavior–geometry paradox: systems can produce human-like responses while relying on a latent structure that differs from human social cognition. We address this gap with a large-scale repeated-elicitation design spanning 16 agents, 65 mental capacities, and 178 replications (185,120 judgments), followed by dimensional analyses and benchmark comparisons with established human mind-perception frameworks. We report three findings. First, model judgments recover a recognizable human-like backbone of mind ascription. Second, this backbone is systematically compressed, with weaker separation among capacities and reduced structural contrast. Third, key dimensions are reconfigured under broader probes, including shifts in how affective, moral-mental, and reality-interaction capacities are partitioned. Together, these results show that behavioral alignment can coexist with representational divergence. The findings support extending safety and alignment evaluation beyond response-level performance to include audits of output-inferred representational geometry, with implications for governance, deployment in persuasive contexts, and theory-building at the intersection of AI and social cognition. Social science/Science, technology and society Business and commerce/Information systems and information technology Full Text Additional Declarations There is NO Competing Interest. Supplementary Files SI18032026Natureanonymised.docx Supporting Information for Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence 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-9161370","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":609111729,"identity":"a927954a-8a84-41aa-8844-e9bd43ad5f26","order_by":0,"name":"DIDIER Grimaldi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYDCCwwwGB8AMZiB+wCYhBxFmI1ZLApuEMWEtBxgMEJwENobEBkJa+I4zbzzwcQ+D3fZ27sQPCWUW6RtuNz9g+FB2GKcWycNsBQdnPGNInnOYd7NEwjmJ3A13jhkwzjiHW4vBYR4gOsCQLMHMu0EisQ2o5UYOAzNvGwEtfyBaNv8Aakk3AGn5S0gLMAzsgFq2gWxJAGthxKMF7JeeAxIJIC0WQL8YzgT65WDPuXScWvjOH9784ccBG3sJ/rObb3woq5Pnu9388MGPMmucWqBAAhodYDYosogA9kjaiVE/CkbBKBgFIwkAAHGMWYXJ6cpDAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-1027-1176","institution":"Polytechnic University of Catalonia","correspondingAuthor":true,"prefix":"","firstName":"DIDIER","middleName":"","lastName":"Grimaldi","suffix":""},{"id":609111730,"identity":"aa208520-0208-4f0f-97c7-20f884b7fcaa","order_by":1,"name":"Carlos Carrasco-Farré","email":"","orcid":"","institution":"Toulouse Business School","correspondingAuthor":false,"prefix":"","firstName":"Carlos","middleName":"","lastName":"Carrasco-Farré","suffix":""}],"badges":[],"createdAt":"2026-03-18 16:03:56","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9161370/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9161370/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106094950,"identity":"c7b2d967-8222-41dd-b91a-191997b5c864","added_by":"auto","created_at":"2026-04-03 11:43:43","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1076757,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript18032026Natureanonymised.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9161370/v1_covered_c1584ef5-ee1b-49c3-9190-25fc2b0eb4bd.pdf"},{"id":105399961,"identity":"139fc949-49a2-41a7-b6be-11c7d1b72cb3","added_by":"auto","created_at":"2026-03-25 15:12:08","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":468644,"visible":true,"origin":"","legend":"Supporting Information for Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence","description":"","filename":"SI18032026Natureanonymised.docx","url":"https://assets-eu.researchsquare.com/files/rs-9161370/v1/7a367fb4734d245fbc206d8c.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"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":"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":"","lastPublishedDoi":"10.21203/rs.3.rs-9161370/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9161370/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"As LLMs increasingly mediate decisions, recommendations, and persuasive communication, a central question is whether their mind-attribution structure in output space is human-like or only behaviorally similar. This creates a behavior–geometry paradox: systems can produce human-like responses while relying on a latent structure that differs from human social cognition. We address this gap with a large-scale repeated-elicitation design spanning 16 agents, 65 mental capacities, and 178 replications (185,120 judgments), followed by dimensional analyses and benchmark comparisons with established human mind-perception frameworks. We report three findings. First, model judgments recover a recognizable human-like backbone of mind ascription. Second, this backbone is systematically compressed, with weaker separation among capacities and reduced structural contrast. Third, key dimensions are reconfigured under broader probes, including shifts in how affective, moral-mental, and reality-interaction capacities are partitioned. Together, these results show that behavioral alignment can coexist with representational divergence. The findings support extending safety and alignment evaluation beyond response-level performance to include audits of output-inferred representational geometry, with implications for governance, deployment in persuasive contexts, and theory-building at the intersection of AI and social cognition.","manuscriptTitle":"Machine Theory of Mind and How Large Language Models Mimic Human Mind Perception but Mask Representational Divergence","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-25 15:11:39","doi":"10.21203/rs.3.rs-9161370/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":"dd94b08a-9815-4fd8-af49-080ff16be6e9","owner":[],"postedDate":"March 25th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":64815670,"name":"Social science/Science, technology and society"},{"id":64815671,"name":"Business and commerce/Information systems and information technology"}],"tags":[],"updatedAt":"2026-04-03T05:10:43+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-25 15:11:39","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9161370","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9161370","identity":"rs-9161370","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}