{"id":863034,"date":"2022-11-22T08:22:05","date_gmt":"2022-11-22T16:22:05","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/"},"modified":"2025-07-28T12:51:47","modified_gmt":"2025-07-28T19:51:47","slug":"language-for-learning-and-interaction","status":"publish","type":"msr-group","link":"https:\/\/www.microsoft.com\/en-us\/research\/theme\/language-for-learning-and-interaction\/","title":{"rendered":"Language for Learning and Interaction | Montr\u00e9al"},"content":{"rendered":"<section class=\"mb-3 moray-highlight\">\n\t<div class=\"card-img-overlay mx-lg-0\">\n\t\t<div class=\"card-background  has-background- card-background--full-bleed\">\n\t\t\t<img loading=\"lazy\" decoding=\"async\" width=\"2560\" height=\"960\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-scaled.jpg\" class=\"attachment-full size-full\" alt=\"Montreal theme header - Language for learning and interaction\" style=\"object-position: 74% 50%\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-scaled.jpg 2560w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-300x113.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-1024x384.jpg 1024w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-768x288.jpg 768w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-1536x576.jpg 1536w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-2048x768.jpg 2048w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-1920x720.jpg 1920w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-1600x600.jpg 1600w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/1920x720_language_for_learning_Montreal_theme_banner_no_logo-240x90.jpg 240w\" sizes=\"auto, (max-width: 2560px) 100vw, 2560px\" \/>\t\t<\/div>\n\t\t<!-- Foreground -->\n\t\t<div class=\"card-foreground d-flex mt-md-n5 my-lg-5 px-g px-lg-0\">\n\t\t\t<!-- Container -->\n\t\t\t<div class=\"container d-flex mt-md-n5 my-lg-5 align-self-center\">\n\t\t\t\t<!-- Card wrapper -->\n\t\t\t\t<div class=\"w-100 w-lg-col-5\">\n\t\t\t\t\t<!-- Card -->\n\t\t\t\t\t<div class=\"card material-md-card py-5 px-md-5\">\n\t\t\t\t\t\t<div class=\"card-body \">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/lab\/microsoft-research-montreal\/\" class=\"icon-link icon-link--reverse mb-2\" data-bi-cN=\"Return to Microsoft Research Lab \u2013 Montr\u00e9al\">\n\t\t\t\t\t\t\t\t\t<span class=\"c-glyph glyph-chevron-left\" aria-hidden=\"true\"><\/span>\n\t\t\t\t\t\t\t\t\tReturn to Microsoft Research Lab \u2013 Montr\u00e9al\t\t\t\t\t\t\t\t<\/a>\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n<h2 class=\"wp-block-heading\" id=\"language-for-learning-and-interaction-montreal\">Language for Learning and Interaction | Montr\u00e9al<\/h2>\n\n\n\n<p>Building interactive language agents that make sense of the world&#8217;s information<\/p>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n<p><em><strong>Building interactive language agents that make sense of the world&#8217;s information, to communicate with and empower people.<\/strong><\/em><\/p>\n\n\n\n<p>This research theme explores how AI agents can use language as a tool to learn from and interact with the world. We envision AI companions that we can teach through conversation and demonstration, that help us in our daily lives.<\/p>\n\n\n\n<p>Language is an abstract space for reasoning, in which modalities\/domains\/tasks can be translated one to another. It is also a medium for vast stores of information. How can agents harness the structure, compositionality, and associations of language to understand and predict a complex world? How can they leverage the knowledge we have built up and stored in writing over the millennia? And how can they use language to share what they\u2019re thinking with us (and each other)? These are some of the questions that animate our research.<\/p>\n\n\n\n<p>Our interdisciplinary work combines methods from natural language processing (NLP), including large pretrained language models, representation learning (especially self- and un-supervised), reinforcement learning, and metalearning.<\/p>\n\n\n\n<p>Methodologically, we take two complementary approaches: Frameworks and Modelling Innovations.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>In <strong>Frameworks<\/strong>, we build learning environments with sequential, dynamic, multi-modal interactions. These enable us to embody agents and ground language in controlled worlds.<\/li>\n\n\n\n<li>In <strong>Modelling Innovations<\/strong>, we explore what architectures, inductive biases, memory mechanisms, and representations best empower agents to use natural language &#8212; in all its generalizing, abstract, and specific complexity. See the Projects tab for more detail.<\/li>\n<\/ul>\n\n\n\n<div class=\"wp-block-media-text has-vertical-margin-small  has-vertical-padding-none  is-stacked-on-mobile is-vertically-aligned-center is-style-border\" style=\"grid-template-columns:30% auto\"><figure class=\"wp-block-media-text__media\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"576\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-1024x576.png\" alt=\"Wordplay language game screen\" class=\"wp-image-867882 size-full\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-1024x576.png 1024w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-300x169.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-768x432.png 768w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-1066x600.png 1066w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-655x368.png 655w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-343x193.png 343w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-240x135.png 240w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-640x360.png 640w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-960x540.png 960w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788-1280x720.png 1280w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/08\/Wordplay-screen_1400x788.png 1400w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure><div class=\"wp-block-media-text__content\">\n<h3 class=\"wp-block-heading\" id=\"workshops-competitions\">Workshops & competitions<\/h3>\n\n\n\n<p>With industry and academic partners, our team co-organizes several research workshops and competitions. This includes the regularly recurring Wordplay workshops, which bring together researchers across a breadth of fields and showcase the interactive narrative community\u2019s challenges, learning environments, and models to the wider NLP\/RL\/Gaming communities.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/wordplay-workshop.github.io\/\" target=\"_blank\" rel=\"noopener noreferrer\">The 3rd Wordplay Workshop: When Language Meets Games<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ NAACL 2022<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.cs.mcgill.ca\/~pparth2\/nilli_workshop_2022\/\" target=\"_blank\" rel=\"noopener noreferrer\">NILLI: Novel Ideas in Learning-to-Learn through Interaction<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ EMNLP 2022<\/li>\n<\/ul>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\">\n<div class=\"wp-block-button is-style-outline is-style-outline--1\"><a data-bi-type=\"button\" class=\"wp-block-button__link wp-element-button\">See past events<\/a><\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"our-collaborators\">Our collaborators<\/h2>\n\n\n\n<p>In addition to the organizations below, we also work closely with other Microsoft teams, including <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/semantic-machines\/\">Semantic Machines<\/a>.<\/p>\n\n\n\n<div class=\"wp-block-columns are-vertically-aligned-top is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-vertically-aligned-top is-layout-flow wp-block-column-is-layout-flow\">\n<figure class=\"wp-block-image aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"300\" height=\"200\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/AI2-logo_300x200.png\" alt=\"Wordplay project - AI2 logo\" class=\"wp-image-862992\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/AI2-logo_300x200.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/AI2-logo_300x200-240x160.png 240w\" sizes=\"auto, (max-width: 300px) 100vw, 300px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-vertically-aligned-top is-layout-flow wp-block-column-is-layout-flow\">\n<figure class=\"wp-block-image aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"300\" height=\"200\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/INRIA-logo_300x200.jpg\" alt=\"INRIA logo\" class=\"wp-image-862995\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/INRIA-logo_300x200.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/INRIA-logo_300x200-240x160.jpg 240w\" sizes=\"auto, (max-width: 300px) 100vw, 300px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-vertically-aligned-top is-layout-flow wp-block-column-is-layout-flow\">\n<figure class=\"wp-block-image aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"797\" height=\"400\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/11\/Logo_Mila_horizontal.png\" alt=\"Mila logo\" class=\"wp-image-627729\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/11\/Logo_Mila_horizontal.png 797w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/11\/Logo_Mila_horizontal-300x151.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/11\/Logo_Mila_horizontal-768x385.png 768w\" sizes=\"auto, (max-width: 797px) 100vw, 797px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-vertically-aligned-top is-layout-flow wp-block-column-is-layout-flow\">\n<figure class=\"wp-block-image aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"300\" height=\"200\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/UofArizona-stack-logo_300x200.png\" alt=\"Wordplay project - University of Arizona logo\" class=\"wp-image-862989\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/UofArizona-stack-logo_300x200.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2022\/07\/UofArizona-stack-logo_300x200-240x160.png 240w\" sizes=\"auto, (max-width: 300px) 100vw, 300px\" \/><\/figure>\n<\/div>\n<\/div>\n\n\n\n\n\n<h2 class=\"wp-block-heading\" id=\"workshops\">Workshops<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/wordplay-workshop.github.io\/\" target=\"_blank\" rel=\"noopener noreferrer\">The 3rd Wordplay Workshop: When Language Meets Games<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ NAACL 2022<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.cs.mcgill.ca\/~pparth2\/nilli_workshop_2022\/\" target=\"_blank\" rel=\"noopener noreferrer\">NILLI: Novel Ideas in Learning-to-Learn through Interaction<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ EMNLP 2022<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.cs.mcgill.ca\/~pparth2\/nilli_workshop\/\" target=\"_blank\" rel=\"noopener noreferrer\">NILLI: Novel Ideas in Learning-to-Learn through Interaction<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ EMNLP 2021<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/slideslive.com\/icml-2020\/1st-workshop-on-language-in-reinforcement-learning-larel\" target=\"_blank\" rel=\"noopener noreferrer\">1st Workshop on Language in Reinforcement Learning (LaReL)<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ ICML 2020<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/kbrl.github.io\/\" target=\"_blank\" rel=\"noopener noreferrer\">KBRL: Knowledge Based Reinforcement Learning<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ IJCAI 2020<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/wordplay-workshop.github.io\/\" target=\"_blank\" rel=\"noopener noreferrer\">The 2nd Wordplay Workshop: When Language Meets Games<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ NeurIPS 2020<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/nips.cc\/Conferences\/2018\/Schedule?showEvent=10938\" target=\"_blank\" rel=\"noopener noreferrer\">The 1st Wordplay Workshop: Reinforcement and Language Learning in Text-based Games<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> @ NeurIPS 2018<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"competitions\">Competitions<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/first-textworld-problems-microsoft-research-montreals-latest-ai-competition-is-really-cooking\/\" target=\"_blank\" rel=\"noreferrer noopener\">First TextWorld Problems: A Reinforcement and Language Learning Challenge<\/a><\/li>\n\n\n\n<li><a href=\"Aligning Text and Embodied Environments for Interactive Learning\" target=\"_blank\" rel=\"noreferrer noopener\">ALFWorld: Aligning Text and Embodied Environments for Interactive Learning<\/a><\/li>\n<\/ul>\n\n\n","protected":false},"excerpt":{"rendered":"<p>Building interactive language agents that make sense of the world&#8217;s information Building interactive language agents that make sense of the world&#8217;s information, to communicate with and empower people. This research theme explores how AI agents can use language as a tool to learn from and interact with the world. We envision AI companions that we [&hellip;]<\/p>\n","protected":false},"featured_media":863577,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_group_start":"","footnotes":""},"research-area":[13556,13545],"msr-group-type":[243688],"msr-locale":[268875],"msr-impact-theme":[],"class_list":["post-863034","msr-group","type-msr-group","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-research-area-human-language-technologies","msr-group-type-theme","msr-locale-en_us"],"msr_group_start":"","msr_detailed_description":"","msr_further_details":"","msr_hero_images":[],"msr_research_lab":[437514],"related-researchers":[{"type":"user_nicename","display_name":"Lucas Caccia","user_id":43647,"people_section":"Section name 0","alias":"lpagecaccia"},{"type":"user_nicename","display_name":"Marc-Alexandre C\u00f4t\u00e9","user_id":37197,"people_section":"Section name 0","alias":"macote"},{"type":"user_nicename","display_name":"Minseon Kim","user_id":43620,"people_section":"Section name 0","alias":"minseonkim"},{"type":"user_nicename","display_name":"Darya Moldavskaya","user_id":43569,"people_section":"Section name 0","alias":"dmoldavskaya"},{"type":"user_nicename","display_name":"Matheus Pereira","user_id":42417,"people_section":"Section name 0","alias":"matpereira"},{"type":"user_nicename","display_name":"Zhengyan Shi","user_id":43864,"people_section":"Section name 0","alias":"zhengyanshi"},{"type":"user_nicename","display_name":"Alessandro Sordoni","user_id":37230,"people_section":"Section name 0","alias":"alsordon"},{"type":"user_nicename","display_name":"Eric Yuan","user_id":37167,"people_section":"Section name 0","alias":"eryua"}],"related-publications":[493214,493226,558498,580978,606396,629274,643860,697849,782797,853254,862314,862320,1135961],"related-downloads":[],"related-videos":[573483,575664],"related-projects":[442191],"related-events":[],"related-opportunities":[],"related-posts":[494219,555504,568296,603186,709156,720673,1136154],"tab-content":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/863034","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-group"}],"version-history":[{"count":27,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/863034\/revisions"}],"predecessor-version":[{"id":1158038,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/863034\/revisions\/1158038"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/863577"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=863034"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=863034"},{"taxonomy":"msr-group-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-group-type?post=863034"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=863034"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=863034"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}