[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-140239-105":59,"doc-detail-140239-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","imitating-interactive-intelligence-21-jan-2021","Imitating Interactive Intelligence - 21 Jan 2021","","The study investigates how to design artificial agents that interact naturally with humans by using a simplified 3D virtual environment. The work addresses core AI challenges: visual perception and goal-directed physical control, grounded language understanding and generation, and multi-agent social interaction. Because training while interacting with humans is impractical, it approximates humans using another learned agent and applies ideas from inverse reinforcement learning to reduce behavioral gaps. It also introduces human-centric behavioral tests and evaluation models aligned with human judgment, enabling scalable assessment of new agent models.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/imitating-interactive-intelligence-21-jan-2021/140239/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/imitating-interactive-intelligence-21-jan-2021/140239.png","ImageObject",300,407,{"name":92,"@type":93},"Mia  ","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-09-17","2026-08-24",true,{"@type":102,"interactionType":103,"userInteractionCount":19},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"Why does the paper use a 3D virtual environment instead of physical robots?","Question",{"text":112,"@type":113},"The virtual domain enables integrated research on perception, control, and language while avoiding technical difficulties of robotic hardware, serving as an ideal testing ground for algorithms and evaluations.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How do the authors approximate human interaction during training?",{"text":117,"@type":113},"They approximate the human role with another learned agent and use inverse reinforcement learning ideas to reduce disparities between human-human and agent-agent interactive behavior.",{"name":119,"@type":110,"acceptedAnswer":120},"How is agent behavior evaluated in the proposed framework?",{"text":121,"@type":113},"The paper develops behavioral tests, including evaluations by humans who watch videos of agents or interact directly. It also trains evaluation models whose ratings agree well with human judgment for scalable evaluation.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},140239,1787571117,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":19,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":129,"read_time":144},687207024478,"https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd","Imitating Interactive Intelligence  \narXiv :2012 .05672v2 [ cs .LG] 21 Jan 2021  \nInteractive Agents Group*  \nDeepMind  \nAbstract  \nA common vision from science fiction is that robots will one day inhabit our physical spaces, sense the world as we do, assist our physical labours, and communicate with us through natural language. Here we study how to design artificial agents that can interact naturally with humans using the simplification of a virtual environment. This setting nevertheless integrates a number of the central challenges of artificial intelligence (AI) research: complex visual perception and goal-directed physical control, grounded language comprehension and production, and multi-agent social interaction. To build agents that can robustly interact with humans, we would ideally train them while they interact with humans. However, this is presently impractical. Therefore, we approximate the role of the human with another learned agent, and use ideas from inverse reinforcement learning to reduce the disparities between human-human and agent-agent interactive behaviour. Rigorously evaluating our agents poses a great challenge, so we develop a variety of behavioural tests, including evaluation by humans who watch videos of agents or interact directly with them. These evaluations convincingly demonstrate that interactive training and auxiliary losses improve agent behaviour beyond what is achieved by supervised learning of actions alone. Further, we demonstrate that agent capabilities generalise beyond literal experiences in the dataset. Finally, we train evaluation models whose ratings of agents agree well with human judgement, thus permitting the evaluation of new agent models without additional effort. Taken together, our results in this virtual environment provide evidence that large-scale human behavioural imitation is a promising tool to create intelligent, interactive agents, and the challenge of reliably evaluating such agents is possible to surmount. See videos for an overview of the manuscript, training time-lapse, and human-agent interactions.  \n* See Section 6 for Authors & Contributions.  \n1 Introduction  \nHumans are an interactive species. We interact with the physical world and with one another. We often attribute our evolved social and linguistic complexity to our intelligence, but this inverts the story: the shaping forces of large-group interactions selected for these capacities (Dunbar, 1993), and these capacities are much of the material of our intelligence. To build artificial intelligence capable of human-like thinking, we therefore must not only grapple with how humans think in the abstract, but also with how humans behave as physical agents in the world and as communicative agents in groups. Our study of how to create artificial agents that interact with humans therefore unifies artificial intelligence with the study of natural human intelligence and behaviour.  \nThis work initiates a research program whose goal is to build embodied artificial agents that can perceive and manipulate the world, understand and produce language, and react capably when given general requests and instructions by humans. Such a holistic research program is consonant with recent calls for more integrated study of the “situated” use of language (McClelland et al., 2019 ; Lake and Murphy, 2020) . Progress towards this goal could greatly expand the scope and naturalness of human-computer interaction (Winograd, 1972 ; Card et al., 1983 ; Branwen, 2018) to the point that interacting with a computer or a robot would be much like interacting with another human being – through shared attention, gesture, demonstration, and dialogue (Tomasello, 2010 ; Winograd, 1972) .  \nOur research program shares much the same spirit as recent work aimed to teach virtual or physical robots to follow instructions provided in natural language (Hermann et al., 2017 ; Lynch and Sermanet, 2020) but attempts to go beyond it by emphasising the interactive and ","cbCaifMOQ6BZSXAa","https://ap.wps.com/l/cbCaifMOQ6BZSXAa","pdf",12636357,96,"English","# Introduction\n## The Virtual Environment\n## The Playroom","[{\"question\":\"Why does the paper use a 3D virtual environment instead of physical robots?\",\"answer\":\"The virtual domain enables integrated research on perception, control, and language while avoiding technical difficulties of robotic hardware, serving as an ideal testing ground for algorithms and evaluations.\"},{\"question\":\"How do the authors approximate human interaction during training?\",\"answer\":\"They approximate the human role with another learned agent and use inverse reinforcement learning ideas to reduce disparities between human-human and agent-agent interactive behavior.\"},{\"question\":\"How is agent behavior evaluated in the proposed framework?\",\"answer\":\"The paper develops behavioral tests, including evaluations by humans who watch videos of agents or interact directly. It also trains evaluation models whose ratings agree well with human judgment for scalable evaluation.\"}]","Imitating Interactive Intelligence - 21 Jan 2021 | PDF",242]