{"id":69,"date":"2025-08-26T20:57:47","date_gmt":"2025-08-26T20:57:47","guid":{"rendered":"https:\/\/biconnect.psicologia.unimib.it\/?page_id=69"},"modified":"2025-09-04T11:04:37","modified_gmt":"2025-09-04T11:04:37","slug":"ai-pedagogy","status":"publish","type":"page","link":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/ai-pedagogy\/","title":{"rendered":"AI Pedagogy"},"content":{"rendered":"\n<p class=\"has-medium-font-size\"><strong>Overview<\/strong><\/p>\n\n\n\n<p>Our project on AI Pedagogy investigates how <strong>teaching and learning strategies<\/strong>\u2014long studied in psychology and education\u2014can be translated into <strong>generative AI systems, especially large language models (LLMs)<\/strong>. At the same time, it asks how these models can serve as <strong>experimental mirrors<\/strong> to refine our understanding of human learning. Rather than treating knowledge acquisition as passive data absorption, we explore <strong>dialogue, scaffolding, and exploration<\/strong>: the very processes through which humans learn best.<\/p>\n\n\n\n<p class=\"has-medium-font-size\"><strong>What we study<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Synthetic learning environments<\/strong>: We design structured, fictional ontologies\u2014artificial but coherent \u201ctoy worlds\u201d\u2014to study learning without the bias of prior knowledge. These environments let us probe how LLMs acquire concepts from scratch.<\/li>\n\n\n\n<li><strong>Teacher\u2013learner dialogues with LLMs<\/strong>: Within these worlds, a <strong>teacher<\/strong> (an LLM with access to the ontology) interacts with a <strong>learner<\/strong> (an LLM or a human who starts na\u00efve). Through structured exchanges, the teacher provides explanations, asks guiding questions, or responds to the learner\u2019s curiosity. We test different strategies inspired by educational psychology: top\u2013down and bottom\u2013up exposition, learner-driven inquiry, teacher-guided questioning, and mixed forms of dialogue.<\/li>\n\n\n\n<li><strong>Humans and LLMs, side by side<\/strong>: LLM simulations allow us to measure which strategies lead to faster and more robust knowledge acquisition. We then bring the same approaches to <strong>human experiments<\/strong>, to compare where LLMs behave like people and where they diverge. Such divergences highlight <strong>human-specific constraints<\/strong> (like working memory, inductive biases, or reliance on social cues), while convergences point to <strong>general principles of pedagogy<\/strong> that may hold across both humans and machines.<\/li>\n\n\n\n<li><strong>Iterative discovery cycle<\/strong>: The project follows a cycle: formalise pedagogical strategies \u2192 test with LLMs \u2192 refine \u2192 test with humans \u2192 interpret differences \u2192 feed back into theory. This process allows us to move from broad pedagogical descriptions to <strong>mechanistic accounts<\/strong> of how knowledge is built through interaction. This translates into turning educational ideas into measurable AI protocols, testing them, comparing humans and machines, and using the insights to sharpen theories of how knowledge is built through interaction.<\/li>\n<\/ul>\n\n\n\n<p class=\"has-medium-font-size\"><strong>Why it matters<\/strong><\/p>\n\n\n\n<p>By embedding pedagogy into LLMs, we aim to produce agents that are not only more adaptive and transparent as learners or teachers, but also provide a <strong>scientific lens on human cognition<\/strong>. Generative AI becomes a kind of <em>microscope for learning<\/em>: making abstract theories of education testable, measurable, and comparable across humans and machines.<\/p>\n\n\n\n<p class=\"has-medium-font-size\"><strong>Methods at a glance<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Environment \u2192 Synthetic ontologies for structured, bias-free learning;<\/li>\n\n\n\n<li>Learner \u2192 Na\u00efve LLM or human participant;<\/li>\n\n\n\n<li>Teacher \u2192 LLM providing explanations and guidance or human participant;<\/li>\n\n\n\n<li>Dialogue \u2192 Pedagogical strategies from psychology (scaffolding, inquiry, exposition);<\/li>\n\n\n\n<li>Evaluation \u2192 Efficiency of knowledge acquisition, quality of generalisation, and human\u2013AI comparisons.<\/li>\n<\/ul>\n\n\n\n<p class=\"has-medium-font-size\"><strong>Project Members<\/strong><\/p>\n\n\n\n<p>Dr. Sabrina Patania, PhD<br>sabrina.patania@unimib.it<\/p>\n\n\n\n<p>Dr. Luca Annese<br>luca.annese1@unimib.it<\/p>\n\n\n\n<p>Dr. Cansu Koyuturk<br>c.koyutuerk@campus.unimib.it<\/p>\n\n\n\n<p>Prof. Dimitri Ognibene<br>dimitri.ognibene@unimib.it<br><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading has-medium-font-size\"><strong>References<\/strong><\/h2>\n\n\n\n<p>Patania, S., Annese, L., Koyuturk, C., Ruggeri, A., &amp; Ognibene, D. (2025). <strong>AI Pedagogy: Dialogic Social Learning for Artificial Agents<\/strong>. <em>Proceedings of the International Conference on Social Robotics (ICSR 2025)<\/em>.<\/p>\n\n\n\n<p>Patania, S., Masiero, E., Brini, L., Piskovskyi, V., Ognibene, D., Donabauer, G., &amp; Kruschwitz, U. (2024). <strong>Large Language Models as an active Bayesian filter: information acquisition and integration<\/strong>. In\u00a0<em>Proceedings of the 28th Workshop on the Semantics and Pragmatics of Dialogue, September, 11-12, 2024, University of Trento<\/em>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Overview Our project on AI Pedagogy investigates how teaching and learning strategies\u2014long studied in psychology and education\u2014can be translated into generative AI systems, especially large language models (LLMs). At the same time, it asks how these models can serve as experimental mirrors to refine our understanding of human learning. Rather than treating knowledge acquisition as &#8230; <a title=\"AI Pedagogy\" class=\"read-more\" href=\"https:\/\/biconnect.psicologia.unimib.it\/index.php\/ai-pedagogy\/\" aria-label=\"Read more about AI Pedagogy\">Read more<\/a><\/p>\n","protected":false},"author":3,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"footnotes":""},"class_list":["post-69","page","type-page","status-publish"],"acf":[],"_links":{"self":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/pages\/69","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/comments?post=69"}],"version-history":[{"count":7,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/pages\/69\/revisions"}],"predecessor-version":[{"id":766,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/pages\/69\/revisions\/766"}],"wp:attachment":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/media?parent=69"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}