Off-Policy Asymptotic and Adaptive Maximum Entropy Deep Reinforcement Learning

Off-Policy Asymptotic and Adaptive Maximum Entropy Deep Reinforcement Learning | 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 Off-Policy Asymptotic and Adaptive Maximum Entropy Deep Reinforcement Learning Huihui Zhang, Xu Han This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4351146/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 13 Oct, 2024 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted 11 You are reading this latest preprint version Abstract Maximum entropy deep reinforcement learning has shown great promise in tackling various challenging continuous tasks. By incorporating the maximum entropy framework, the goal is to introduce more randomness in action selection and improve the training process. However, there exists a tradeoff between efficiency and stability, especially when dealing with large-scale tasks with high state and action dimensions.In certain situations, it becomes necessary to constrain the temperature hyperparameter of the maximum entropy term to prevent instability, which can hinder convergence. In this study, we propose an algorithm that combines adaptive and asymptotic maximum entropy with actor-critic random policies.Specifically, we introduce a state-dependent adaptive temperature to accelerate the training process and include an additional term involving asymptotic maximum entropy to ensure stable convergence. These components are combined with the selected critic value to serve as the target Q-value and the surrogate objective in the policy evaluation and improvement steps.The adaptive and asymptotic maximum entropy algorithm demonstrates robust adaptation to the efficiency-stability tradeoff, providing increased exploration and flexibility to address saddle point problems. We evaluate our method on various Gym tasks, and the results indicate that our proposed algorithms outperform several baselines in the domain of continuous control. offline continuous control reinforcement learning behavior regularization Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 13 Oct, 2024 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted Editorial decision: Revision requested 07 Jun, 2024 Reviews received at journal 03 Jun, 2024 Reviews received at journal 02 Jun, 2024 Reviews received at journal 22 May, 2024 Reviewers agreed at journal 08 May, 2024 Reviewers agreed at journal 07 May, 2024 Reviewers agreed at journal 05 May, 2024 Reviewers invited by journal 05 May, 2024 Editor assigned by journal 05 May, 2024 Submission checks completed at journal 02 May, 2024 First submitted to journal 30 Apr, 2024 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-4351146","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":299878716,"identity":"0fa03faa-79e9-4af4-a52e-282a2c537132","order_by":0,"name":"Huihui Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEUlEQVRIiWNgGAWjYBACAwbGNhCZAGQzPkgwYJMDiR54QKQWZoMPBXzGYC0JeLUwsIFokBo2yRkf5BIbYFxcwFwiue0xT0FdHv/s9gvSPAZm6fPDDj8E2mInp9uAXYvljMR2Yx4DtmKJO2cKgIy03I230wyAWpKNzQ7gcNiNxDag4TyJDTdyEpJ5DI7lbpydANJyIHEbfi0SifOBWg7zGPxPN5yd/oEYLQaJG26kH2ycYcCWIC+dQ8CWMw/bJOcYJCRuvJHDzPDBgM1wg3ROwYEEAzx+OZ7+TOLNn7rEeTfSn/9I+MMmLz87ffOHDxV2cri0IAEeA4ghYJUGBJWDAPsDMCXfQJTqUTAKRsEoGEEAABw9ZkpviMs4AAAAAElFTkSuQmCC","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Huihui","middleName":"","lastName":"Zhang","suffix":""},{"id":299878718,"identity":"3b2089ff-f3dc-471b-bdde-ab2ea236af98","order_by":1,"name":"Xu Han","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Xu","middleName":"","lastName":"Han","suffix":""}],"badges":[],"createdAt":"2024-04-30 19:14:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4351146/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4351146/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s13042-024-02399-7","type":"published","date":"2024-10-13T15:58:05+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":66597290,"identity":"e84048da-df5e-4434-9a8a-e59540938994","added_by":"auto","created_at":"2024-10-14 16:09:14","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":946657,"visible":true,"origin":"","legend":"","description":"","filename":"AAACIJMLC.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4351146/v1_covered_1ac8e6f6-ded6-4aa4-9773-8cb7488db4e1.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Off-Policy Asymptotic and Adaptive Maximum Entropy Deep Reinforcement Learning","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":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"offline, continuous control, reinforcement learning, behavior regularization","lastPublishedDoi":"10.21203/rs.3.rs-4351146/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4351146/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Maximum entropy deep reinforcement learning has shown great promise in tackling various challenging continuous tasks. By incorporating the maximum entropy framework, the goal is to introduce more randomness in action selection and improve the training process. However, there exists a tradeoff between efficiency and stability, especially when dealing with large-scale tasks with high state and action dimensions.In certain situations, it becomes necessary to constrain the temperature hyperparameter of the maximum entropy term to prevent instability, which can hinder convergence. In this study, we propose an algorithm that combines adaptive and asymptotic maximum entropy with actor-critic random policies.Specifically, we introduce a state-dependent adaptive temperature to accelerate the training process and include an additional term involving asymptotic maximum entropy to ensure stable convergence. These components are combined with the selected critic value to serve as the target Q-value and the surrogate objective in the policy evaluation and improvement steps.The adaptive and asymptotic maximum entropy algorithm demonstrates robust adaptation to the efficiency-stability tradeoff, providing increased exploration and flexibility to address saddle point problems. We evaluate our method on various Gym tasks, and the results indicate that our proposed algorithms outperform several baselines in the domain of continuous control.","manuscriptTitle":"Off-Policy Asymptotic and Adaptive Maximum Entropy Deep Reinforcement Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-08 18:27:18","doi":"10.21203/rs.3.rs-4351146/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-08T02:51:24+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-03T17:14:53+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-03T00:50:08+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-22T07:45:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"276343148687655632615649309018869852263","date":"2024-05-08T06:23:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"83218228459354899519472438699265035771","date":"2024-05-07T04:49:33+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"116443326924649135416145971466008923469","date":"2024-05-05T17:42:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-05-05T13:14:15+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-05-05T11:00:31+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-05-02T12:43:55+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Machine Learning and Cybernetics","date":"2024-04-30T19:13:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"561ed201-4ade-4994-85c7-b677b406d7e4","owner":[],"postedDate":"May 8th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-10-14T16:03:43+00:00","versionOfRecord":{"articleIdentity":"rs-4351146","link":"https://doi.org/10.1007/s13042-024-02399-7","journal":{"identity":"international-journal-of-machine-learning-and-cybernetics","isVorOnly":false,"title":"International Journal of Machine Learning and Cybernetics"},"publishedOn":"2024-10-13 15:58:05","publishedOnDateReadable":"October 13th, 2024"},"versionCreatedAt":"2024-05-08 18:27:18","video":"","vorDoi":"10.1007/s13042-024-02399-7","vorDoiUrl":"https://doi.org/10.1007/s13042-024-02399-7","workflowStages":[]},"version":"v1","identity":"rs-4351146","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4351146","identity":"rs-4351146","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}