Model Predictive Task Sampling for Efficient and Robust Adaptation

Model Predictive Task Sampling for Efficient and Robust Adaptation | 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 Model Predictive Task Sampling for Efficient and Robust Adaptation Qi (Cheems) Wang, Zehao Xiao, Yixiu Mao, Yun Qu, Jiayi Shen, Yiqin Lv, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6700167/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 Foundation models have revolutionized general-purpose problem-solving, offering rapidtask adaptation through pretraining, meta-training, and finetuning. Recent crucial advances in these paradigms reveal the importance of challenging taskprioritized sampling to enhance adaptation robustness under distribution shifts. However, ranking task difficulties over iteration as a preliminary step typically requiresexhaustive task evaluation, which is practically unaffordable in computation and data-annotation. This study provides a novel perspective to illuminate the possibility of leveraging thedual importance of adaptation robustness and learning efficiency, particularly inscenarios where task evaluation is risky or costly, such as iterative agent-environmentinteractions for robotic policy evaluation or computationally intensive inference steps forfinetuning foundation models. Firstly, we introduce Model Predictive Task Sampling (MPTS), a framework that bridgesthe task space and adaptation risk landscape, providing a theoretical foundation forrobust active task sampling. MPTS employs a generative model to characterize the episodic optimization process andpredicts task-specific adaptation risk via posterior inference. The resulting risk learner amortizes the costly evaluation of task adaptationperformance and provably approximates task difficulty rankings. MPTS seamlesslyintegrates into zero-shot, few-shot, and supervised finetuning settings. Empirically, we conduct extensive experiments in pattern recognition using foundationmodels and sequential decision-making. Our results demonstrate that MPTS significantly enhances adaptation robustness for tailor out-of-distribution (OOD) tasks and improves learning efficiency compared to state-of-the-art (SOTA) methods. The code is available at the project site https://github.com/thu-rllab/MPTS. Artificial Intelligence and Machine 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-6700167","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":458826508,"identity":"d5b4a5ef-8096-4f99-9b6d-0d3759df2a06","order_by":0,"name":"Qi (Cheems) Wang","email":"data:image/png;base64,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","orcid":"","institution":"Tsinghua University","correspondingAuthor":true,"prefix":"","firstName":"Qi","middleName":"(Cheems)","lastName":"Wang","suffix":""},{"id":458828124,"identity":"5e8e52b8-4770-415d-9df5-700c7b10ae86","order_by":1,"name":"Zehao Xiao","email":"","orcid":"","institution":"University of Amsterdam","correspondingAuthor":false,"prefix":"","firstName":"Zehao","middleName":"","lastName":"Xiao","suffix":""},{"id":458828125,"identity":"7e7a59d6-3c07-4451-9cc1-290029089dcb","order_by":2,"name":"Yixiu Mao","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Yixiu","middleName":"","lastName":"Mao","suffix":""},{"id":458828126,"identity":"4a2f875e-4a9d-44cb-a6dc-81b9cabab043","order_by":3,"name":"Yun Qu","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Yun","middleName":"","lastName":"Qu","suffix":""},{"id":458828127,"identity":"75493a86-81d6-4d19-ad3b-e315dc21cf10","order_by":4,"name":"Jiayi Shen","email":"","orcid":"","institution":"University of Amsterdam","correspondingAuthor":false,"prefix":"","firstName":"Jiayi","middleName":"","lastName":"Shen","suffix":""},{"id":458828128,"identity":"a41ca19c-eda9-403d-a645-b9a435afe2da","order_by":5,"name":"Yiqin Lv","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Yiqin","middleName":"","lastName":"Lv","suffix":""},{"id":458828129,"identity":"0f3bd90d-f871-4316-80a6-6b041c3a7f4a","order_by":6,"name":"Xiangyang Ji","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Xiangyang","middleName":"","lastName":"Ji","suffix":""}],"badges":[],"createdAt":"2025-05-19 15:00:13","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":true,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":true},"doi":"10.21203/rs.3.rs-6700167/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6700167/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83102950,"identity":"b31eaff0-bed1-47ac-a5af-d6b16389db73","added_by":"auto","created_at":"2025-05-20 05:32:06","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8038184,"visible":true,"origin":"","legend":"","description":"","filename":"MPTSArxiv.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6700167/v1_covered_650c8e14-9a6c-43ae-830d-492f423da48c.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eModel Predictive Task Sampling for Efficient and Robust Adaptation\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Tsinghua University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"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-6700167/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6700167/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFoundation models have revolutionized general-purpose problem-solving, offering rapidtask adaptation through pretraining, meta-training, and finetuning. Recent crucial advances in these paradigms reveal the importance of challenging taskprioritized sampling to enhance adaptation robustness under distribution shifts. However, ranking task difficulties over iteration as a preliminary step typically requiresexhaustive task evaluation, which is practically unaffordable in computation and data-annotation. This study provides a novel perspective to illuminate the possibility of leveraging thedual importance of adaptation robustness and learning efficiency, particularly inscenarios where task evaluation is risky or costly, such as iterative agent-environmentinteractions for robotic policy evaluation or computationally intensive inference steps forfinetuning foundation models. Firstly, we introduce Model Predictive Task Sampling (MPTS), a framework that bridgesthe task space and adaptation risk landscape, providing a theoretical foundation forrobust active task sampling. MPTS employs a generative model to characterize the episodic optimization process andpredicts task-specific adaptation risk via posterior inference. The resulting risk learner amortizes the costly evaluation of task adaptationperformance and provably approximates task difficulty rankings. MPTS seamlesslyintegrates into zero-shot, few-shot, and supervised finetuning settings. Empirically, we conduct extensive experiments in pattern recognition using foundationmodels and sequential decision-making. Our results demonstrate that MPTS significantly enhances adaptation robustness for tailor out-of-distribution (OOD) tasks and improves learning efficiency compared to state-of-the-art (SOTA) methods. The code is available at the project site https://github.com/thu-rllab/MPTS.\u003c/p\u003e","manuscriptTitle":"Model Predictive Task Sampling for Efficient and Robust Adaptation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-20 05:07:50","doi":"10.21203/rs.3.rs-6700167/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":"ca9a4064-1829-4032-8f9a-4e920bd58807","owner":[],"postedDate":"May 20th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48742690,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2025-06-03T13:21:00+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-20 05:07:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6700167","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6700167","identity":"rs-6700167","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}