{"paper_id":"15c5c7ff-b53d-4a9c-af0c-db0f11601fad","body_text":"IMPACTX: Improving Model Performance by Appropriately predicting CorrecT eXplanations | 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 IMPACTX: Improving Model Performance by Appropriately predicting CorrecT eXplanations Andrea Apicella, Francesco Isgrò, Salvatore Giugliano, Andrea Pollastro, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8060140/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 16 Apr, 2026 Read the published version in Artificial Intelligence Review → Version 1 posted 10 You are reading this latest preprint version Abstract The eXplainable Artificial Intelligence (XAI) research predominantly concentrates to provide explainations about AI model decisions, especially Deep Learning (DL) models. However, there is a growing interest in using XAI techniques to automatically improve the performance of the AI systems themselves. This paper proposes IMPACTX, a novel approach that leverages XAI as a fully automated attention mechanism, without requiring external knowledge or human feedback. Experimental results show that IMPACTX has improved performance respect to the standalone ML model by integrating an attention mechanism based an XAI method outputs during the model training. Furthermore, IMPACTX directly provides proper feature attribution maps for the model's decisions, without relying on external XAI methods during the inference process. Our proposal is evaluated using three widely recognized DL models (EfficientNet-B2, MobileNet, and LeNet-5) along with three standard image datasets: CIFAR-10, CIFAR-100, and STL-10. The results show that IMPACTX consistently improves the performance of all the inspected DL models across all evaluated datasets, and it directly provides appropriate explanations for its responses. XAI performance improvement deep learning explanations attribution maps Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 16 Apr, 2026 Read the published version in Artificial Intelligence Review → Version 1 posted Reviewers agreed at journal 23 Nov, 2025 Reviews received at journal 19 Nov, 2025 Reviewers agreed at journal 19 Nov, 2025 Reviewers agreed at journal 18 Nov, 2025 Reviewers agreed at journal 18 Nov, 2025 Reviewers agreed at journal 18 Nov, 2025 Reviewers invited by journal 18 Nov, 2025 Editor assigned by journal 16 Nov, 2025 Submission checks completed at journal 08 Nov, 2025 First submitted to journal 07 Nov, 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-8060140\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":549456268,\"identity\":\"1af92392-3f47-4c37-9b55-d5236c617e00\",\"order_by\":0,\"name\":\"Andrea Apicella\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIiWNgGAWjYHACAzDiY2dgfMDAIAEUYGwgTgsbMwOzAVRLIyE9BmASqIVNAiqC3xr+2c0bPxcU2MmxMTM/q+bdYWHP38Dc/gCfFok7x4qlZxgkG7Mxs5nd5j0jkTjjACGH3cgxkOYxOJDYxswA1NImkcBASIv8jRzj30At9W3M7N+KgVrs5QlpMbiRYwayJYGNmceMGaiFcQMhLYY30sqsgX4xbGPmKZac2yaRuPEwY+MMfFrkbiRvvl3wx06en71944e3bXX2csfbH3zApwUEmPFyidEyCkbBKBgFowANAACCk0F5rJPeGQAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"University of Salerno\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Andrea\",\"middleName\":\"\",\"lastName\":\"Apicella\",\"suffix\":\"\"},{\"id\":549456269,\"identity\":\"ac365bac-c0eb-4ca0-8fc2-205a1889093d\",\"order_by\":1,\"name\":\"Francesco Isgrò\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Naples Federico II\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Francesco\",\"middleName\":\"\",\"lastName\":\"Isgrò\",\"suffix\":\"\"},{\"id\":549456270,\"identity\":\"f5a1270f-bf2d-4a81-a240-38f2620ed9eb\",\"order_by\":2,\"name\":\"Salvatore Giugliano\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Istituto di Calcolo e Reti ad Alte Prestazioni (ICAR) Centro Nazionale delle Ricerche (CNR), Centro Nazionale delle Ricerche (CNR)\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Salvatore\",\"middleName\":\"\",\"lastName\":\"Giugliano\",\"suffix\":\"\"},{\"id\":549456271,\"identity\":\"00f0f61e-c7f9-472e-a5e0-4a23eb4fc9d3\",\"order_by\":3,\"name\":\"Andrea Pollastro\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Naples Federico II\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Andrea\",\"middleName\":\"\",\"lastName\":\"Pollastro\",\"suffix\":\"\"},{\"id\":549456272,\"identity\":\"e14e4fd8-44da-4947-b12e-b9b59a0d5628\",\"order_by\":4,\"name\":\"Roberto Prevete\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Naples Federico II\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Roberto\",\"middleName\":\"\",\"lastName\":\"Prevete\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-11-07 21:08:15\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-8060140/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-8060140/v1\",\"draftVersion\":[],\"editorialEvents\":[{\"content\":\"https://doi.org/10.1007/s10462-026-11564-z\",\"type\":\"published\",\"date\":\"2026-04-16T15:59:53+00:00\"}],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":96860686,\"identity\":\"72d5f52a-2204-4344-8994-21f79b30b1fb\",\"added_by\":\"auto\",\"created_at\":\"2025-11-26 22:12:21\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":4037938,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"IMPACTX.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8060140/v1/180ff1694b57727cd77fdc2e.pdf\"},{\"id\":96860688,\"identity\":\"4abe122b-c881-4fa6-b634-ff5a3973b030\",\"added_by\":\"auto\",\"created_at\":\"2025-11-26 22:12:22\",\"extension\":\"json\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":5944,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"ec1c5edb5a4b4775986992a5064512ac.json\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8060140/v1/7c8bd15de92f5cb60b5e9cc4.json\"},{\"id\":107351422,\"identity\":\"9b031bef-5b2a-4f07-a0a9-d9c48c68fd44\",\"added_by\":\"auto\",\"created_at\":\"2026-04-20 16:10:51\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":992592,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"IMPACTX.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8060140/v1_covered_79035b16-2598-447e-8c7f-869c3b84fb62.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"IMPACTX: Improving Model Performance by Appropriately predicting CorrecT eXplanations\",\"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\":\"info@researchsquare.com\",\"identity\":\"artificial-intelligence-review\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"aire\",\"sideBox\":\"Learn more about [Artificial Intelligence Review](http://link.springer.com/journal/10462)\",\"snPcode\":\"10462\",\"submissionUrl\":\"https://submission.nature.com/new-submission/10462/3\",\"title\":\"Artificial Intelligence Review\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"Springer Hybrid\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false},\"keywords\":\" XAI, performance improvement, deep learning, explanations, attribution maps\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-8060140/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-8060140/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eThe eXplainable Artificial Intelligence (XAI) research predominantly concentrates to provide explainations about AI model decisions, especially Deep Learning (DL) models. \\u0026nbsp;However, there is a growing interest in using XAI techniques to automatically improve the performance of the AI systems themselves.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eThis paper proposes IMPACTX, a novel approach that leverages XAI as a fully automated attention mechanism, without requiring external knowledge or human feedback. Experimental results show that IMPACTX has improved performance respect to the standalone ML model by integrating an attention mechanism based an XAI method outputs during the model training. Furthermore, IMPACTX directly provides proper feature attribution maps for the model's decisions, without relying on external XAI methods during the inference process.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eOur proposal is evaluated using three widely recognized DL models (EfficientNet-B2, MobileNet, and LeNet-5) along with three standard image datasets: CIFAR-10, CIFAR-100, and STL-10. The results show that IMPACTX consistently improves the performance of all the inspected DL models across all evaluated datasets, and it directly provides appropriate explanations for its responses.\\u003c/p\\u003e\",\"manuscriptTitle\":\"IMPACTX: Improving Model Performance by Appropriately predicting CorrecT eXplanations\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-11-26 22:12:16\",\"doi\":\"10.21203/rs.3.rs-8060140/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0},{\"type\":\"reviewerAgreed\",\"content\":\"298817107487307683900097351433062772577\",\"date\":\"2025-11-23T17:25:26+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2025-11-19T08:33:22+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"287384655329713075695151991673066907553\",\"date\":\"2025-11-19T08:05:11+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"243279574218668427081674549115034200660\",\"date\":\"2025-11-18T16:26:51+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"334631806226681411095838193762727434903\",\"date\":\"2025-11-18T12:27:53+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"142354394360871352782877394194364167614\",\"date\":\"2025-11-18T11:57:19+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewersInvited\",\"content\":\"\",\"date\":\"2025-11-18T11:54:26+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorAssigned\",\"content\":\"\",\"date\":\"2025-11-16T12:15:54+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"checksComplete\",\"content\":\"\",\"date\":\"2025-11-08T06:59:03+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"submitted\",\"content\":\"Artificial Intelligence Review\",\"date\":\"2025-11-07T21:05:44+00:00\",\"index\":\"\",\"fulltext\":\"\"}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"artificial-intelligence-review\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"aire\",\"sideBox\":\"Learn more about [Artificial Intelligence Review](http://link.springer.com/journal/10462)\",\"snPcode\":\"10462\",\"submissionUrl\":\"https://submission.nature.com/new-submission/10462/3\",\"title\":\"Artificial Intelligence Review\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"Springer Hybrid\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false}}],\"origin\":\"\",\"ownerIdentity\":\"e978948f-f2f6-4715-84a1-88b236a4af77\",\"owner\":[],\"postedDate\":\"November 26th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"published-in-journal\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2026-04-20T16:08:24+00:00\",\"versionOfRecord\":{\"articleIdentity\":\"rs-8060140\",\"link\":\"https://doi.org/10.1007/s10462-026-11564-z\",\"journal\":{\"identity\":\"artificial-intelligence-review\",\"isVorOnly\":false,\"title\":\"Artificial Intelligence Review\"},\"publishedOn\":\"2026-04-16 15:59:53\",\"publishedOnDateReadable\":\"April 16th, 2026\"},\"versionCreatedAt\":\"2025-11-26 22:12:16\",\"video\":\"\",\"vorDoi\":\"10.1007/s10462-026-11564-z\",\"vorDoiUrl\":\"https://doi.org/10.1007/s10462-026-11564-z\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-8060140\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-8060140\",\"identity\":\"rs-8060140\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}