Retrieval-augmented generation: a hybrid approach to assessing retrieved documents similarity, LLM confidence, and system stability

Retrieval-augmented generation: a hybrid approach to assessing retrieved documents similarity, LLM confidence, and system stability | 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 Short Report Retrieval-augmented generation: a hybrid approach to assessing retrieved documents similarity, LLM confidence, and system stability Alexander Plesovskikh This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6741053/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 retrieval-augmented generation approach relies on retrieving relevant documents to enhance large language model prompts and improve model outputs. However, existing metrics like cosine similarity, precision@k, and recall@k, to name just a few, fail to account for the confidence and stability of retrieval and generation. We propose a novel approach, retrieval confidence score, and its extension, asymptotic retrieval confidence score, which combines semantic similarity, large language model confidence, and stability across multiple generations. Asymptotic retrieval confidence score potentially provides a robust approach for evaluating retrieval-augmented generation systems, possibly suggesting a better solution for combining results across retrieval, generation, and evaluation stages. Large language model Metrics Retrieval-augmented generation Full Text Additional Declarations No competing interests reported. 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-6741053","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Short Report","associatedPublications":[],"authors":[{"id":461690232,"identity":"a1c6f194-57a3-40fa-bd67-5480d9641eee","order_by":0,"name":"Alexander Plesovskikh","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBklEQVRIie3RMUvDQBTA8XccOF3sailJvsLJQafgZ7lQiIsI4tLBIeUgXVK7NiD6FTJ1vvDALqlzJxVcHQIuHQ4q6STq1bp1uP/ylvvBexyAy3WYEb0djCgKQyApsL/NF1L/jwClJNuDdMYTrRsTxQ85KnF1H/ml9rBp4DlOLeSkfpJVkSVxuRypQTFPRKmPk+4Mrq0EVhccvRQF7xCF3hzjUrM+ZSCtJHx552gMinDakrtNS8SH2UH4inGEIwygXcxLdUt4D3aQ0/pSVpMsCXhdKcEeB6JA1u/mXAobCRbL6nVtIhbm5289dnPm3y5y0ayH0ree//MX6HZh6/vfiMvlcrm+9QnmAl/cW98olgAAAABJRU5ErkJggg==","orcid":"","institution":"Higher School of Economics","correspondingAuthor":true,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Plesovskikh","suffix":""}],"badges":[],"createdAt":"2025-05-24 22:28:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6741053/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6741053/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96453858,"identity":"23f912c7-c420-49d3-9476-29736e26ff79","added_by":"auto","created_at":"2025-11-21 10:01:55","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":444904,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6741053/v1_covered_c905327a-7811-46e9-a06f-9ed9991a009e.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Retrieval-augmented generation: a hybrid approach to assessing retrieved documents similarity, LLM confidence, and system stability","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":"Large language model, Metrics, Retrieval-augmented generation","lastPublishedDoi":"10.21203/rs.3.rs-6741053/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6741053/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe retrieval-augmented generation approach relies on retrieving relevant documents to enhance large language model prompts and improve model outputs. However, existing metrics like cosine similarity, precision@k, and recall@k, to name just a few, fail to account for the confidence and stability of retrieval and generation. We propose a novel approach, retrieval confidence score, and its extension, asymptotic retrieval confidence score, which combines semantic similarity, large language model confidence, and stability across multiple generations. Asymptotic retrieval confidence score potentially provides a robust approach for evaluating retrieval-augmented generation systems, possibly suggesting a better solution for combining results across retrieval, generation, and evaluation stages.\u003c/p\u003e","manuscriptTitle":"Retrieval-augmented generation: a hybrid approach to assessing retrieved documents similarity, LLM confidence, and system stability","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-28 04:33:55","doi":"10.21203/rs.3.rs-6741053/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":"5f922e5c-6aaf-4a50-b036-6c068ac91083","owner":[],"postedDate":"May 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-11-20T14:23:56+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-28 04:33:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6741053","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6741053","identity":"rs-6741053","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}