[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84174-en":3,"doc-seo-84174-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84174,1374391974468,"Eden","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Learning social norms enhances compatibility in dynamic human–AI coordination","Humans coordinate with others in dynamic interactions using implicit, hard-to-quantify social norms that act as shared tacit expectations among agents. As AI agents, including large language models, join daily interactions, they often fail to coordinate effectively, considerately, and naturally. The work hypothesizes the gap comes from aligning behavior to demonstrations without explicitly quantifying the underlying norms. Using a pedestrian–vehicle platform and 3,456 collected interactions, three norm principles are identified: outcome predictability, value alignment, and advantage awareness.","Learning social norms enhances compatibility in dynamic  \nhuman–AI coordination  \nYi Yang1, 6, Siyuan Liu 1, 6, Xin Gao 1, Huamu Sun 1, Chao Liu3, 4, 5, Qing Zhou 1, Bingbing Nie 1, 2, *  \nAbstract  \nHumans continuously coordinate with others in dynamic interactions, often through implicit, hard-to-quantify social norms that act as shared tacit expectations among interacting agents 1,2 . As AI agents, including large language models (LLMs), become embedded in daily life, they increasingly participate in such interactions and reshape social interaction structures3,4 . Yet they often fail to coordinate with humans in an effective, considerate, and natural manner5-8. We hypothesize that this gap arises because existing approaches align model behavior with human demonstrations without explicitly quantifying the underlying norms that generate such behavior. We selected pedestrianvehicle interaction as a representative dynamic interaction and developed a simplified experimental platform that captures its key interactive features. From 3,456 dynamic human interactions collected via this platform, we identified three principles underlying human social norms: outcome predictability, value alignment, and advantage awareness.  \nIncorporating these principles into AI agents significantly improves human–AI coordination. In the closed-loop interaction task with humans, the social-norm-informed LLM achieved a nearly fourfold higher total score than the baseline strategy and outperformed human–human interactions by 43% . These findings indicate that formalizing tacit social norms into explicit, quantifiable principles can enable AI agents to achieve mutually beneficial coordination in dynamic interactions, supporting their more natural integration into human society.  \n1 Introduction  \nDynamic interactions between humans and artificial intelligence (AI) are increasingly embedded in realworld physical and social environments4,9,10 . In such settings, humans and AI agents engage in continuous, history-dependent, and mutually adaptive decision-making processes, where outcomes emerge through ongoing interactions between agents 11. Representative scenarios include autonomous vehicles negotiating right-of-way with pedestrians12, service robots navigating dense crowds 13, and AI agents engaging with humans in decision-making tasks such as logistics coordination and emergency response 14. While many aspects of these interactions are regulated by explicit operational constraints, such as traffic regulations and physical safety limits15-17, the constraints alone are insufficient to fully define successful coordination. A key limitation is that core processes of human coordination, such as anticipation, reciprocity, and mutual adaptation, are governed by implicit social norms 1,2 . If AI agents interacting with humans lack such normaware reasoning, inappropriate or even absurd consequences could follow. In terms of interaction quality, violations of social norms by AI can erode trust, increase cognitive friction, and induce frustration in human  \n1School of Vehicle and Mobility, Tsinghua University, Beijing, China. 2State Key Laboratory of Intelligent Green Vehicle and Mobility, Tsinghua University, Beijing 100084, China. 3State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China. 4Beijing Key Laboratory of Safe AI and Superalignment, Beijing, China. 5Beijing Institute of AI Safety and Governance, Beijing, China. 6These authors contributed equally: Yi Yang and Siyuan Liu. * Correspondence [to: ](to: nbb@tsinghua.edu.cn)[nbb@tsinghua.edu.cn](to: nbb@tsinghua.edu.cn)  \nusers18-21. In terms of physical safety, they introduce unpredictability into human–AI encounters that may escalate into hazardous outcomes22. For instance, pedestrians interacting with autonomous vehicles that lack social norms may experience heightened uncertainty and adopt defensive or assertive mane","cbCait29kx8kL6ZO","https://ap.wps.com/l/cbCait29kx8kL6ZO","pdf",2485401,1,44,"English","en",105,"# Abstract\n# Introduction","[{\"question\":\"Why do AI agents struggle with natural human–AI coordination in dynamic settings?\",\"answer\":\"They often lack explicit, norm-aware reasoning. Existing methods align behavior to demonstrations or optimize rewards without quantifying the social norms that generate human coordination.\"},{\"question\":\"How was the problem studied in the paper?\",\"answer\":\"The paper uses pedestrian–vehicle interaction as a representative dynamic scenario. A simplified experimental platform collects 3,456 dynamic human interactions for analysis.\"},{\"question\":\"What principles were extracted from human social norms, and what effect do they have?\",\"answer\":\"Three principles were identified: outcome predictability, value alignment, and advantage awareness. Incorporating them into AI agents improves closed-loop coordination, yielding a much higher total score and outperforming human–human interactions by 43%.\"}]",1784193637,111,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"learning-social-norms-enhances-compatibility-in-dynamic-humanai-coordination","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/learning-social-norms-enhances-compatibility-in-dynamic-humanai-coordination/84174/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why do AI agents struggle with natural human–AI coordination in dynamic settings?","Question",{"text":75,"@type":76},"They often lack explicit, norm-aware reasoning. Existing methods align behavior to demonstrations or optimize rewards without quantifying the social norms that generate human coordination.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How was the problem studied in the paper?",{"text":80,"@type":76},"The paper uses pedestrian–vehicle interaction as a representative dynamic scenario. A simplified experimental platform collects 3,456 dynamic human interactions for analysis.",{"name":82,"@type":73,"acceptedAnswer":83},"What principles were extracted from human social norms, and what effect do they have?",{"text":84,"@type":76},"Three principles were identified: outcome predictability, value alignment, and advantage awareness. Incorporating them into AI agents improves closed-loop coordination, yielding a much higher total score and outperforming human–human interactions by 43%.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]