{"id":1003089,"date":"2024-01-30T06:27:35","date_gmt":"2024-01-30T14:27:35","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-project&#038;p=1003089"},"modified":"2024-01-30T07:06:17","modified_gmt":"2024-01-30T15:06:17","slug":"ai-frontiers-explorations","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/ai-frontiers-explorations\/","title":{"rendered":"AI Frontiers: Explorations"},"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=\"1921\" height=\"720\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720.png\" class=\"attachment-full size-full\" alt=\"AI and Microsoft Research header - abstract neural network pattern on dark spectrum background\" style=\"\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720.png 1921w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-300x112.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-1024x384.png 1024w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-768x288.png 768w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-1536x576.png 1536w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-1600x600.png 1600w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2023\/03\/AI_Microsoft_Research_Header_1920x720-240x90.png 240w\" sizes=\"(max-width: 1921px) 100vw, 1921px\" \/>\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 \">\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\/?p=992148&post_type=msr-research-lab&preview=1&_ppp=253bdc9b27\" class=\"icon-link icon-link--reverse mb-2\" data-bi-cN=\"AI Frontiers\">\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\tAI Frontiers\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<h1 class=\"wp-block-heading\" id=\"explorations\">Explorations<\/h1>\n\n\n\n<p><\/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>We are actively exploring a range of projects that we believe hold significant potential not only for Microsoft but also for the broader society.<\/p>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"large-action-behavioral-models\">Large Action & Behavioral Models<\/h2>\n\n\n\n<p class=\"h5\">Games, Robotics, Self-driving OS<\/p>\n\n\n\n<p>Large Action & Behavioral Models (LAM) are AI foundational models that can predict how to act on human environments and interfaces, like LLM are capable to interpret human language and generate coherent text. We develop LAMs that will be capable to perform complex tasks in a multitude of scenarios, like how to play a videogame, how a to move in the real world, and how to use an Operating System (OS).<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"transformer-enhancements\">Transformer Enhancements<\/h2>\n\n\n\n<p class=\"h5\">EMA (Exponential moving average), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/publication\/the-truth-is-in-there-improving-reasoning-in-language-models-with-layer-selective-rank-reduction\/\">LASER<\/a>, Learning from Text feedback, feedback driven learning in a RAG<\/p>\n\n\n\n<p>We develop techniques that can dramatically improve the performance or reduce the costs of Transformer-based Large Language Models (LLMs) by applying simple interventions after training requiring minimal or zero parameters or data.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"foundational-models-for-latent-state-and-latent-dynamics-in-llm\">Foundational Models for latent state and latent dynamics in LLM<\/h2>\n\n\n\n<p>A latent state or a latent dynamic can improve planning, exploration, and credit assignment of AI agents by keeping task-relevant information and discarding distractions. We develop techniques that autonomously discover such latent representations directly from real-world signals.<\/p>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n","protected":false},"excerpt":{"rendered":"<p>We are actively exploring a range of projects that we believe hold significant potential not only for Microsoft but also for the broader society. Games, Robotics, Self-driving OS Large Action & Behavioral Models (LAM) are AI foundational models that can predict how to act on human environments and interfaces, like LLM are capable to interpret [&hellip;]<\/p>\n","protected":false},"featured_media":992742,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"research-area":[13556],"msr-locale":[268875],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1003089","msr-project","type-msr-project","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-locale-en_us","msr-archive-status-active"],"msr_project_start":"","related-publications":[1054431],"related-downloads":[],"related-videos":[],"related-groups":[],"related-events":[],"related-opportunities":[],"related-posts":[],"related-articles":[],"tab-content":[],"slides":[],"related-researchers":[],"msr_research_lab":[992148],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1003089"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":16,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1003089\/revisions"}],"predecessor-version":[{"id":1003299,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1003089\/revisions\/1003299"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/992742"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1003089"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1003089"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1003089"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1003089"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1003089"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}