Tomas Pfister

<!DOCTYPE html> <html lang="en"> <head> <meta charset="utf-8" /> <link rel="canonical" href="http://research.google/people/105803/" /><meta property="og:title" content="Tomas Pfister"><meta property="og:url" content="http://research.google/people/105803/"><meta property="og:image" content="https://storage.googleapis.com/gweb-research2023-media/images/Open_Graph.width-800.format-jpeg.jpg"><meta property="og:image:secure_url" content="https://storage.googleapis.com/gweb-research2023-media/images/Open_Graph.width-800.format-jpeg.jpg"><meta property="og:type" content="Website"> <title>Tomas Pfister</title> <meta name="viewport" content="width=device-width, initial-scale=1 viewport-fit=cover"/> <link rel="icon" type="image/png" href="/gr/static/assets/favicon.ico"> <link rel="preconnect" href="https://fonts.googleapis.com"> <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> <link rel="preload" href="https://fonts.googleapis.com/css2?family=Product+Sans&family=Google+Sans+Display:ital@0;1&family=Google+Sans:ital,wght@0,400;0,500;0,700;1,400;1,500;1,700&family=Google+Sans+Text:ital,wght@0,400;0,500;0,700;1,400;1,500;1,700&display=swap" as="style"> <link rel="stylesheet" href="https://fonts.googleapis.com/css2?family=Product+Sans&family=Google+Sans+Display:ital@0;1&family=Google+Sans:ital,wght@0,400;0,500;0,700;1,400;1,500;1,700&family=Google+Sans+Text:ital,wght@0,400;0,500;0,700;1,400;1,500;1,700&display=swap"> <link href="https://fonts.googleapis.com/css2?family=Roboto+Mono:wght@400;700&display=swap" rel="stylesheet"> <link href="https://www.gstatic.com/glue/cookienotificationbar/cookienotificationbar.min.css" rel="stylesheet" /> <link href="https://www.gstatic.com/glue/v27_1/glue-material.min.css" rel="stylesheet"> <link rel="stylesheet" type="text/css" href="/gr/static/css/googleresearch.css?id=0c26ea1fed8bdd0324f9f4fad1f6a470">  <script>(function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start': new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0], j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src= 'https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f); })(window,document,'script','dataLayer','GTM-K8QBZ7Q'); </script>  </head> <body class=" js-google-tag-wrapper" data-gt-page-path="http://research.google/people/105803/" data-env="production">  <noscript><iframe src="https://www.googletagmanager.com/ns.html?id=GTM-K8QBZ7Q" height="0" width="0" style="display:none;visibility:hidden"></iframe></noscript>  <header class="global-header glue-header glue-header--single not-glue"> <a href="#page-content" class="glue-header__skip-content">Jump to Content</a> <div class="glue-header__bar glue-header__bar--mobile not-glue"> <div class="glue-header__tier not-glue">  <div class="glue-header__container"> <div class="glue-header__lock-up">  <div class="glue-header__hamburger"> <button class="glue-header__drawer-toggle-btn" aria-label="Open the navigation drawer"> <svg class="glue-icon glue-icon--24px" role="presentation" aria-hidden="true"> <use href="/gr/static/assets/icons/glue-icons.svg#menu"></use> </svg> </button> </div> <div class="glue-header__logo"> <a class="glue-header__logo-link" href="/" title="Google Research">  <div class="glue-header__logo-container"> <svg role="presentation" aria-hidden="true" alt='Google' class="glue-icon glue-icon glue-header__logo-svg"> <use href="/gr/static/assets/icons/glue-icons.svg#google-color-logo"></use> </svg> </div> <span class="glue-header__logo--product">Research</span> </a> </div> </div> </div> </div> </div> <div class="glue-header__bar glue-header__bar--desktop glue-header__drawer"> <div class="glue-header__tier">  <div class="glue-header__container"> <div class="glue-header__lock-up"> <div class="glue-header__logo"> <a class="glue-header__logo-link" href="/" title="Google Research">  <div class="glue-header__logo-container"> <svg role="presentation" aria-hidden="true" alt='Google' class="glue-icon glue-icon glue-header__logo-svg not-glue --dark-logo"> <use href="/gr/static/assets/icons/glue-icons.svg#google-solid-logo"></use> </svg> <svg role="presentation" aria-hidden="true" alt='Google' class="glue-icon glue-icon glue-header__logo-svg --light-logo"> <use href="/gr/static/assets/icons/glue-icons.svg#google-color-logo"></use> </svg> </div> <span class="glue-header__logo--product">Research</span> </a> </div> </div> </div>  <div class="glue-header__container glue-header__container--linkbar"> <nav class="glue-header__link-bar navigation js-gt-global-nav-wrapper"> <ul class="glue-header__list"> <li class="glue-header__item js-sub-nav-parent --parent" data-gt-primary="Who we are" > <button class="glue-header__link js-sub-nav-target" aria-haspopup="true" aria-expanded="false" > <span class=""> Who we are <span class="icon icon--caret"></span> </span> </button> <div class="navigation__sub js-sub-nav" role="menu"> <div class="navigation__sub__container"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Who we are <span class="sr-text">menu</span> </button> <hr/> </div> <div class="block-nav_drawer_columns_content"> <div class="navigation__sub--content" data-gt-secondary="Defining the technology of today and tomorrow."> <div class="navigation__sub__wrapper"> <div class="navigation__sub__heading"> <h2 class="headline-3">Defining the technology of today and tomorrow.</h2> </div> <ul class="navigation__sub__columns"> <li data-gt-secondary="Philosophy"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Philosophy </h2> <p class="navigation__sub__columns__description caption">We strive to create an environment conducive to many different types of research across many different time scales and levels of risk.</p> <a href="http://research.google/philosophy/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Philosophy</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/philosophy/" > Philosophy </a> </div> </li> <li data-gt-secondary="People"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> People </h2> <p class="navigation__sub__columns__description caption">Our researchers drive advancements in computer science through both fundamental and applied research.</p> <a href="http://research.google/people/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our People</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/people/" > People </a> </div> </li> </ul> </div> </div> </div> </div> </div> </li> <li class="glue-header__item js-sub-nav-parent --parent" data-gt-primary="Research areas" > <button class="glue-header__link js-sub-nav-target" aria-haspopup="true" aria-expanded="false" > <span class=""> Research areas <span class="icon icon--caret"></span> </span> </button> <div class="navigation__sub js-sub-nav" role="menu"> <div class="navigation__sub__container"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Research areas <span class="sr-text">menu</span> </button> <hr/> </div> <div class="block-nav_drawer_columns_link_list"> <div class="navigation__sub--list"> <div class="navigation__sub__wrapper"> <ul class="navigation__sub__columns"> <li data-gt-secondary="Research areas"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading">Research areas</h2> <ul> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/" > Explore all research areas </a> </li> </ul> </div> <div class="navigation__sub__columns__mobile"> <button class="glue-header__link js-sub-nav-target" data-panel="nested" role="menuitem" aria-haspopup="true"> Research areas <span class="icon icon--caret"></span> </button> <div class="navigation__nested-sub js-sub-nav-parent"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile" role="menuitem" aria-haspopup="true"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Research areas <span class="sr-text">menu</span> </button> <hr/> </div> <ul> <li role="menuitem"> <a href="http://research.google/research-areas/" class="navigation__sub__columns__mobile__link" > Explore all research areas <span> </span> </a> </li> </ul> </div> </div> </li> <li data-gt-secondary="Foundational ML & Algorithms"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading">Foundational ML & Algorithms</h2> <ul> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/algorithms-and-theory/" > Algorithms & Theory </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/data-management/" > Data Management </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/data-mining-and-modeling/" > Data Mining & Modeling </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/information-retrieval-and-the-web/" > Information Retrieval & the Web </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/machine-intelligence/" > Machine Intelligence </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/machine-perception/" > Machine Perception </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/machine-translation/" > Machine Translation </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/natural-language-processing/" > Natural Language Processing </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/speech-processing/" > Speech Processing </a> </li> </ul> </div> <div class="navigation__sub__columns__mobile"> <button class="glue-header__link js-sub-nav-target" data-panel="nested" role="menuitem" aria-haspopup="true"> Foundational ML & Algorithms <span class="icon icon--caret"></span> </button> <div class="navigation__nested-sub js-sub-nav-parent"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile" role="menuitem" aria-haspopup="true"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Foundational ML & Algorithms <span class="sr-text">menu</span> </button> <hr/> </div> <ul> <li role="menuitem"> <a href="http://research.google/research-areas/algorithms-and-theory/" class="navigation__sub__columns__mobile__link" > Algorithms & Theory <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/data-management/" class="navigation__sub__columns__mobile__link" > Data Management <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/data-mining-and-modeling/" class="navigation__sub__columns__mobile__link" > Data Mining & Modeling <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/information-retrieval-and-the-web/" class="navigation__sub__columns__mobile__link" > Information Retrieval & the Web <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/machine-intelligence/" class="navigation__sub__columns__mobile__link" > Machine Intelligence <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/machine-perception/" class="navigation__sub__columns__mobile__link" > Machine Perception <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/machine-translation/" class="navigation__sub__columns__mobile__link" > Machine Translation <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/natural-language-processing/" class="navigation__sub__columns__mobile__link" > Natural Language Processing <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/speech-processing/" class="navigation__sub__columns__mobile__link" > Speech Processing <span> </span> </a> </li> </ul> </div> </div> </li> <li data-gt-secondary="Computing Systems & Quantum AI"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading">Computing Systems & Quantum AI</h2> <ul> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/distributed-systems-and-parallel-computing/" > Distributed Systems & Parallel Computing </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/hardware-and-architecture/" > Hardware & Architecture </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/mobile-systems/" > Mobile Systems </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/networking/" > Networking </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/quantum-computing/" > Quantum Computing </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/robotics/" > Robotics </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/security-privacy-and-abuse-prevention/" > Security, Privacy, & Abuse Prevention </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/software-engineering/" > Software Engineering </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/software-systems/" > Software Systems </a> </li> </ul> </div> <div class="navigation__sub__columns__mobile"> <button class="glue-header__link js-sub-nav-target" data-panel="nested" role="menuitem" aria-haspopup="true"> Computing Systems & Quantum AI <span class="icon icon--caret"></span> </button> <div class="navigation__nested-sub js-sub-nav-parent"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile" role="menuitem" aria-haspopup="true"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Computing Systems & Quantum AI <span class="sr-text">menu</span> </button> <hr/> </div> <ul> <li role="menuitem"> <a href="http://research.google/research-areas/distributed-systems-and-parallel-computing/" class="navigation__sub__columns__mobile__link" > Distributed Systems & Parallel Computing <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/hardware-and-architecture/" class="navigation__sub__columns__mobile__link" > Hardware & Architecture <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/mobile-systems/" class="navigation__sub__columns__mobile__link" > Mobile Systems <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/networking/" class="navigation__sub__columns__mobile__link" > Networking <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/quantum-computing/" class="navigation__sub__columns__mobile__link" > Quantum Computing <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/robotics/" class="navigation__sub__columns__mobile__link" > Robotics <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/security-privacy-and-abuse-prevention/" class="navigation__sub__columns__mobile__link" > Security, Privacy, & Abuse Prevention <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/software-engineering/" class="navigation__sub__columns__mobile__link" > Software Engineering <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/software-systems/" class="navigation__sub__columns__mobile__link" > Software Systems <span> </span> </a> </li> </ul> </div> </div> </li> <li data-gt-secondary="Science, AI & Society"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading">Science, AI & Society</h2> <ul> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/climate-and-sustainability/" > Climate & Sustainability </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/economics-and-electronic-commerce/" > Economics & Electronic Commerce </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/education-innovation/" > Education Innovation </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/general-science/" > General Science </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/health-bioscience/" > Health & Bioscience </a> </li> <li> <a class="navigation__sub__columns__list-link caption js-drawer-link" href="http://research.google/research-areas/human-computer-interaction-and-visualization/" > Human-Computer Interaction and Visualization </a> </li> </ul> </div> <div class="navigation__sub__columns__mobile"> <button class="glue-header__link js-sub-nav-target" data-panel="nested" role="menuitem" aria-haspopup="true"> Science, AI & Society <span class="icon icon--caret"></span> </button> <div class="navigation__nested-sub js-sub-nav-parent"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile" role="menuitem" aria-haspopup="true"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Science, AI & Society <span class="sr-text">menu</span> </button> <hr/> </div> <ul> <li role="menuitem"> <a href="http://research.google/research-areas/climate-and-sustainability/" class="navigation__sub__columns__mobile__link" > Climate & Sustainability <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/economics-and-electronic-commerce/" class="navigation__sub__columns__mobile__link" > Economics & Electronic Commerce <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/education-innovation/" class="navigation__sub__columns__mobile__link" > Education Innovation <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/general-science/" class="navigation__sub__columns__mobile__link" > General Science <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/health-bioscience/" class="navigation__sub__columns__mobile__link" > Health & Bioscience <span> </span> </a> </li> <li role="menuitem"> <a href="http://research.google/research-areas/human-computer-interaction-and-visualization/" class="navigation__sub__columns__mobile__link" > Human-Computer Interaction and Visualization <span> </span> </a> </li> </ul> </div> </div> </li> </ul> </div> </div></div> </div> </div> </li> <li class="glue-header__item js-sub-nav-parent --parent" data-gt-primary="Our work" > <button class="glue-header__link js-sub-nav-target" aria-haspopup="true" aria-expanded="false" > <span class=""> Our work <span class="icon icon--caret"></span> </span> </button> <div class="navigation__sub js-sub-nav" role="menu"> <div class="navigation__sub__container"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Our work <span class="sr-text">menu</span> </button> <hr/> </div> <div class="block-nav_drawer_columns_content"> <div class="navigation__sub--content" data-gt-secondary=""> <div class="navigation__sub__wrapper"> <ul class="navigation__sub__columns"> <li data-gt-secondary="Projects"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Projects </h2> <p class="navigation__sub__columns__description caption">We regularly open-source projects with the broader research community and apply our developments to Google products.</p> <a href="http://research.google/resources/our-projects/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Projects</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/resources/our-projects/" > Projects </a> </div> </li> <li data-gt-secondary="Publications"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Publications </h2> <p class="navigation__sub__columns__description caption">Publishing our work allows us to share ideas and work collaboratively to advance the field of computer science.</p> <a href="http://research.google/pubs/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Publications</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/pubs/" > Publications </a> </div> </li> <li data-gt-secondary="Resources"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Resources </h2> <p class="navigation__sub__columns__description caption">We make products, tools, and datasets available to everyone with the goal of building a more collaborative ecosystem.</p> <a href="http://research.google/resources/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Resources</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/resources/" > Resources </a> </div> </li> </ul> </div> </div> </div> </div> </div> </li> <li class="glue-header__item js-sub-nav-parent --parent" data-gt-primary="Programs & events" > <button class="glue-header__link js-sub-nav-target" aria-haspopup="true" aria-expanded="false" > <span class=""> Programs & events <span class="icon icon--caret"></span> </span> </button> <div class="navigation__sub js-sub-nav" role="menu"> <div class="navigation__sub__container"> <div class="navigation__sub__mobile-heading"> <button class="glue-header__link js-sub-nav-close-mobile"> <span class="sr-text">Back to</span> <span class="icon icon--caret"></span> Programs & events <span class="sr-text">menu</span> </button> <hr/> </div> <div class="block-nav_drawer_columns_content"> <div class="navigation__sub--content" data-gt-secondary="Shaping the future, together."> <div class="navigation__sub__wrapper"> <div class="navigation__sub__heading"> <h2 class="headline-3">Shaping the future, together.</h2> <a href="http://research.google/programs-and-events/" class="js-drawer-link" > Collaborate with us </a> </div> <ul class="navigation__sub__columns"> <li data-gt-secondary="Student programs"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Student programs </h2> <p class="navigation__sub__columns__description caption">Supporting the next generation of researchers through a wide range of programming.</p> <a href="http://research.google/programs-and-events/student-engagement/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Student programs</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/programs-and-events/student-engagement/" > Student programs </a> </div> </li> <li data-gt-secondary="Faculty programs"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Faculty programs </h2> <p class="navigation__sub__columns__description caption">Participating in the academic research community through meaningful engagement with university faculty.</p> <a href="http://research.google/programs-and-events/faculty-engagement/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Faculty programs</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/programs-and-events/faculty-engagement/" > Faculty programs </a> </div> </li> <li data-gt-secondary="Conferences & events"> <div class="navigation__sub__columns__desktop"> <h2 class="headline-6 navigation__sub__columns__heading"> Conferences & events </h2> <p class="navigation__sub__columns__description caption">Connecting with the broader research community through events is essential for creating progress in every aspect of our work.</p> <a href="http://research.google/conferences-and-events/" class="glue-inline-link js-drawer-link" > <span class="sr-text">Learn more about our Conferences & events</span> <span aria-hidden="true">Learn more</span> </a> </div> <div class="navigation__sub__columns__mobile"> <a class="glue-header__link" href="http://research.google/conferences-and-events/" > Conferences & events </a> </div> </li> </ul> <div class="navigation__sub__cta"> <a class="glue-button glue-button--high-emphasis js-drawer-link" href="http://research.google/programs-and-events/" target="_blank" rel="noreferrer noopener" > Collaborate with us </a> </div> </div> </div> </div> </div> </div> </li> <li class="glue-header__item " data-gt-primary="Careers" > <a class="glue-header__link " href="http://research.google/careers/" > <span class=""> Careers </span> </a> </li> <li class="glue-header__item " data-gt-primary="Blog" > <a class="glue-header__link " href="http://research.google/blog/" > <span class=""> Blog </span> </a> </li> </ul> </nav> </div>  <div class="glue-header__search js-header-search"> <div class="glue-header__search__input"> <div class="search-input " data-type="header"> <input type="search" class="caption --empty-search js-search-bar js-gt-search-input" placeholder="Search"> <button class="search-input__button --search js-gt-search-btn"> <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--18px "> <use href="/gr/static/assets/icons/glue-icons.svg#search"></use> </svg> </button> <button class="search-input__button --clear"> <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--18px "> <use href="/gr/static/assets/icons/glue-icons.svg#close"></use> </svg> </button> </div> </div> <button class="glue-header__search__btn js-header-search-btn"> <svg role="presentation" aria-hidden="true" aria-hidden="true" class="glue-icon glue-icon--24px search"> <use href="/gr/static/assets/icons/glue-icons.svg#search"></use> </svg> <svg role="presentation" aria-hidden="true" aria-hidden="true" class="glue-icon glue-icon--24px close"> <use href="/gr/static/assets/icons/glue-icons.svg#close"></use> </svg> <span class="sr-text js-header-search-sr-text">Search</span> </button> </div> </div> </div> <div class="glue-header__drawer-backdrop"> <div class="glue-header__mobile_close"> <button class="glue-header__drawer-toggle-btn js-mobile-nav-close" aria-label="Close the navigation drawer"> <svg class="glue-icon glue-icon--24px" role="presentation" aria-hidden="true"> <use href="/gr/static/assets/icons/glue-icons.svg#close"></use> </svg> </button> </div> </div> </header> <main id="page-content"> <div class="person-detail-page"> <section class="basic-hero bhoig --theme-dark " data-gt-id="basic_hero" data-gt-component-name=""> <div class="glue-page"> <div class="glue-grid"> <div class="bhoig__image-wrapper glue-grid__col--span-2 glue-grid__col--span-3-md glue-grid__col--span-2-lg"> <div class="bhoig__image-bg" style=""> <picture> <img src="https://storage.googleapis.com/gweb-research2023-media/pubtools/4434.png" alt="Tomas Pfister" class=" basic-hero__person-image "/> </picture> </div> </div> <div class="bhoig__breadcrumb-wrapper glue-grid__col--span-10 glue-grid__col--span-9-md glue-grid__col--span-10-lg"> <nav class="glue-breadcrumbs" aria-label="Breadcrumbs"> <ol class="glue-breadcrumbs__list"> <li class="glue-breadcrumbs__item"> <a class="glue-breadcrumbs__link attribution" href="/">Home</a> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#chevron-right"></use> </svg> </li> <li class="glue-breadcrumbs__item"> <a class="glue-breadcrumbs__link attribution" href="/people/">People</a> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#chevron-right"></use> </svg> </li> </ol> </nav> </div> <h1 class="headline-1 bhoig__headline glue-grid__col--span-10 glue-grid__col--span-9-md glue-grid__col--span-10-lg">Tomas Pfister</h1> <div class="basic-hero__description bhoig__description glue-grid__col--span-10 glue-grid__col--span-9-md glue-grid__col--span-10-lg"> Tomas Pfister is the Head of Cloud AI Research. He came to Google from Apple where he cofounded Apple's central AI research group and published Apple’s first research paper that won the Best Paper Award at CVPR’17. Tomas’ key scientific achievements have been proposing a method to improve the realism of synthetic images; developing the first automated method to detect facial micro-expressions; and inventing a new way for neural networks to exploit spatiotemporal structure. He is currently exploring learning from small amount of labeled data (using techniques such as generative models, few-shot learning, transfer learning) and explainability/interpretability of deep learning models, and is particularly excited about the potential of AI in healthcare & education. His research has laid the foundation for several applications such as Face ID in iPhone X, autonomous driving, human pose estimation, detecting facial micro-expressions & translating sign language. Tomas did his PhD in deep learning with Prof Andrew Zisserman at Oxford University and bachelor’s degree in computer science at Cambridge University. He is the recipient of the Forbes 30 Under 30 award, and has received over 40 research awards, including 3 best paper awards, with numerous publications in top AI research venues. His work has been frequently featured in mainstream media, including Forbes, BusinessInsider & Wired. </div> <div class="bhoig__cta glue-grid__col--span-10 glue-grid__col--span-9-md glue-grid__col--span-10-lg"> </div> </div> </div> </section> <section class="offset-two-up --research-areas"> <div class="glue-page glue-grid"> <div class="offset-two-up__left-col glue-grid__col glue-grid__col--span-4-sm glue-grid__col--span-12-md glue-grid__col--span-3-lg"> <h3 class="offset-two-up__headline headline-3">Research Areas</h3> </div> <div class="glue-grid__col glue-grid__col--span-4-sm glue-grid__col--span-12-md glue-grid__col--span-9-lg"> <ul class="card-stack--small nested-glue-grid-override"> <li class="glue-grid__col glue-grid__col--span-6-md glue-grid__col--span-4-sm"> <a class="glue-card not-glue glue-card--small" href="/research-areas/algorithms-and-theory/"> <div class="glue-card__inner"> <div class="glue-card__content"> <div class="content-wrapper"> <p class="glue-headline body js-gt-item-id"> Algorithms and Theory </p> </div> <div class="glue-card__cta glue-card__cta--arrow"> <span class="glue-button glue-button--low-emphasis glue-button--icon"> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-forward"></use> </svg> </span> </div> </div> </div> </a> </li> <li class="glue-grid__col glue-grid__col--span-6-md glue-grid__col--span-4-sm"> <a class="glue-card not-glue glue-card--small" href="/research-areas/machine-intelligence/"> <div class="glue-card__inner"> <div class="glue-card__content"> <div class="content-wrapper"> <p class="glue-headline body js-gt-item-id"> Machine Intelligence </p> </div> <div class="glue-card__cta glue-card__cta--arrow"> <span class="glue-button glue-button--low-emphasis glue-button--icon"> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-forward"></use> </svg> </span> </div> </div> </div> </a> </li> <li class="glue-grid__col glue-grid__col--span-6-md glue-grid__col--span-4-sm"> <a class="glue-card not-glue glue-card--small" href="/research-areas/machine-perception/"> <div class="glue-card__inner"> <div class="glue-card__content"> <div class="content-wrapper"> <p class="glue-headline body js-gt-item-id"> Machine Perception </p> </div> <div class="glue-card__cta glue-card__cta--arrow"> <span class="glue-button glue-button--low-emphasis glue-button--icon"> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-forward"></use> </svg> </span> </div> </div> </div> </a> </li> <li class="glue-grid__col glue-grid__col--span-6-md glue-grid__col--span-4-sm"> <a class="glue-card not-glue glue-card--small" href="/research-areas/natural-language-processing/"> <div class="glue-card__inner"> <div class="glue-card__content"> <div class="content-wrapper"> <p class="glue-headline body js-gt-item-id"> Natural Language Processing </p> </div> <div class="glue-card__cta glue-card__cta--arrow"> <span class="glue-button glue-button--low-emphasis glue-button--icon"> <svg role="presentation" aria-hidden="true" class="glue-icon "> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-forward"></use> </svg> </span> </div> </div> </div> </a> </li> </ul> </div> </div> </section> <div class="glue-page heading-tabs-wrapper spacer-03-top"> <span class="--as-h4">Authored Publications</span> </div>  <section class="publication-index filtered-list-base hot-swap-base glue-page theme-light spacer-06-top spacer-06-bottom"> <div class="js-configurable-list glue-grid" data-list-id="publications-index-list" data-search-query-param=""> <div class="filter-wrapper"> <div class="glue-filter glue-filter--mws" id="my-filter" data-target-id="publications-index-list" data-chipset-id="publications-index-chipset" data-use-facet-names="True" >  <div class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-12-md">  <button class="glue-button glue-button--low-emphasis glue-button--icon glue-filter__modal-toggle"> See Filters <svg role="presentation" aria-hidden="true" class="glue-icon"> <use href="/gr/static/assets/icons/glue-icons.svg#filter"></use> </svg> </button>  <div class="glue-modal glue-filter__modal" aria-labelledby="modal-label"> <div id="modal-content" class="glue-filter__modal-content"> <output class="glue-filter__results-status glue-visually-hidden" aria-live="polite" aria-atomic="true"> <span class="glue-filter__results-count"></span> results </output> <div class="glue-filter__categories glue-filter__panels glue-expansion-panels js-gt-filter-wrapper" data-glue-expansion-panels-key="filter-mws" data-glue-expansion-panel-expand-tooltip="Expand this filter category" data-glue-expansion-panel-collapse-tooltip="Collapse this filter category">  <div class="glue-expansion-panels__header"> <h2 class="glue-headline caption" id="modal-label"> Filter by: </h2>  <div class="glue-filter__modal-controls glue-filter__filter-controls js-gt-filter-clear"> <button class="glue-button glue-button--low-emphasis glue-filter__reset-all" > <span class="js-gt-item-id">Clear</span> </button> </div> <button class="glue-modal__close-btn" title="Close the modal" tabindex="0" aria-label="Close the modal"> </button> </div>  <div class="glue-expansion-panel filter__first-panel"> <h3 class="glue-expansion-panel__toggle"> <span class="glue-expansion-panel__button" id="filter-toggle-type" data-glue-expansion-panel-toggle-for="filter-panel-type"> <span class="glue-expansion-panel__header-text" id="filter-type-label">Publications</span> <svg role="presentation" aria-hidden="true" class="glue-icon glue-expansion-panel__header-arrow"> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-drop-down"></use> </svg> </span> </h3>  <div class="glue-expansion-panel__content" id="filter-panel-type"> <div> <ul class="glue-filter__category glue-no-bullet" data-glue-filter-category="type" role="list" aria-controls="filter-results" aria-labelledby="filter-type-label"> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-type-google" name="filter-type-google" value="google" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-type-google"> <span class="js-filter-option-label-text">Google</span> <span class="filter-option-label-count">53</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-type-other" name="filter-type-other" value="other" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-type-other"> <span class="js-filter-option-label-text">Other</span> <span class="filter-option-label-count">3</span></label> </div> </div> </li> </ul> </div> </div> </div> <div class="glue-expansion-panel "> <h3 class="glue-expansion-panel__toggle"> <span class="glue-expansion-panel__button" id="filter-toggle-year" data-glue-expansion-panel-toggle-for="filter-panel-year"> <span class="glue-expansion-panel__header-text" id="filter-year-label">Years</span> <svg role="presentation" aria-hidden="true" class="glue-icon glue-expansion-panel__header-arrow"> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-drop-down"></use> </svg> </span> </h3>  <div class="glue-expansion-panel__content" id="filter-panel-year"> <div> <ul class="glue-filter__category glue-no-bullet" data-glue-filter-category="year" role="list" aria-controls="filter-results" aria-labelledby="filter-year-label"> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2024" name="filter-year-2024" value="2024" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2024"> <span class="js-filter-option-label-text">2024</span> <span class="filter-option-label-count">6</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2023" name="filter-year-2023" value="2023" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2023"> <span class="js-filter-option-label-text">2023</span> <span class="filter-option-label-count">17</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2022" name="filter-year-2022" value="2022" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2022"> <span class="js-filter-option-label-text">2022</span> <span class="filter-option-label-count">9</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2021" name="filter-year-2021" value="2021" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2021"> <span class="js-filter-option-label-text">2021</span> <span class="filter-option-label-count">9</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2020" name="filter-year-2020" value="2020" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2020"> <span class="js-filter-option-label-text">2020</span> <span class="filter-option-label-count">9</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2019" name="filter-year-2019" value="2019" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2019"> <span class="js-filter-option-label-text">2019</span> <span class="filter-option-label-count">3</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2017" name="filter-year-2017" value="2017" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2017"> <span class="js-filter-option-label-text">2017</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2016" name="filter-year-2016" value="2016" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2016"> <span class="js-filter-option-label-text">2016</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-year-2015" name="filter-year-2015" value="2015" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-year-2015"> <span class="js-filter-option-label-text">2015</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> </ul> </div> </div> </div> <div class="glue-expansion-panel "> <h3 class="glue-expansion-panel__toggle"> <span class="glue-expansion-panel__button" id="filter-toggle-research_area" data-glue-expansion-panel-toggle-for="filter-panel-research_area"> <span class="glue-expansion-panel__header-text" id="filter-research_area-label">Research Areas</span> <svg role="presentation" aria-hidden="true" class="glue-icon glue-expansion-panel__header-arrow"> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-drop-down"></use> </svg> </span> </h3>  <div class="glue-expansion-panel__content" id="filter-panel-research_area"> <div> <ul class="glue-filter__category glue-no-bullet" data-glue-filter-category="research_area" role="list" aria-controls="filter-results" aria-labelledby="filter-research_area-label"> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-algorithms-and-theory" name="filter-research_area-algorithms-and-theory" value="algorithms-and-theory" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-algorithms-and-theory"> <span class="js-filter-option-label-text">Algorithms and Theory</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-data-management" name="filter-research_area-data-management" value="data-management" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-data-management"> <span class="js-filter-option-label-text">Data Management</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-health-bioscience" name="filter-research_area-health-bioscience" value="health-bioscience" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-health-bioscience"> <span class="js-filter-option-label-text">Health & Bioscience</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-machine-intelligence" name="filter-research_area-machine-intelligence" value="machine-intelligence" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-machine-intelligence"> <span class="js-filter-option-label-text">Machine Intelligence</span> <span class="filter-option-label-count">45</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-machine-perception" name="filter-research_area-machine-perception" value="machine-perception" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-machine-perception"> <span class="js-filter-option-label-text">Machine Perception</span> <span class="filter-option-label-count">19</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-machine-translation" name="filter-research_area-machine-translation" value="machine-translation" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-machine-translation"> <span class="js-filter-option-label-text">Machine Translation</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-natural-language-processing" name="filter-research_area-natural-language-processing" value="natural-language-processing" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-natural-language-processing"> <span class="js-filter-option-label-text">Natural Language Processing</span> <span class="filter-option-label-count">8</span></label> </div> </div> </li> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-research_area-responsible-ai" name="filter-research_area-responsible-ai" value="responsible-ai" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-research_area-responsible-ai"> <span class="js-filter-option-label-text">Responsible AI</span> <span class="filter-option-label-count">1</span></label> </div> </div> </li> </ul> </div> </div> </div> <div class="glue-expansion-panel "> <h3 class="glue-expansion-panel__toggle"> <span class="glue-expansion-panel__button" id="filter-toggle-team" data-glue-expansion-panel-toggle-for="filter-panel-team"> <span class="glue-expansion-panel__header-text" id="filter-team-label">Teams</span> <svg role="presentation" aria-hidden="true" class="glue-icon glue-expansion-panel__header-arrow"> <use href="/gr/static/assets/icons/glue-icons.svg#arrow-drop-down"></use> </svg> </span> </h3>  <div class="glue-expansion-panel__content" id="filter-panel-team"> <div> <ul class="glue-filter__category glue-no-bullet" data-glue-filter-category="team" role="list" aria-controls="filter-results" aria-labelledby="filter-team-label"> <li class="glue-filter__category-item js-filter-item js-gt-filter-item"> <div class="mdc-touch-target-wrapper"> <div class="mdc-form-field"> <div class="mdc-checkbox mdc-checkbox--touch glue-checkbox"> <input type="checkbox" class="mdc-checkbox__native-control" id="filter-team-cloud-ai" name="filter-team-cloud-ai" value="cloud-ai" > <div class="mdc-checkbox__background"> <svg class="mdc-checkbox__checkmark" viewBox="0 0 24 24"> <path class="mdc-checkbox__checkmark-path" fill="none" d="M1.73,12.91 8.1,19.28 22.79,4.59"></path> </svg> <div class="mdc-checkbox__mixedmark"></div> </div> <div class="mdc-checkbox__ripple"></div> </div> <label class="filter-option-label" for="filter-team-cloud-ai"> <span class="js-filter-option-label-text">Cloud AI</span> <span class="filter-option-label-count">39</span></label> </div> </div> </li> </ul> </div> </div> </div> </div> </div> </div> </div>  <div class="glue-filter__results mock-results" data-glue-filter-categories="type year research_area team" id="filter-results" > <ul class="glue-cards glue-grid"> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="google"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="other"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2015"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2016"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2017"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2019"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2020"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2021"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2022"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2023"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="2024"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="algorithms-and-theory"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="data-management"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="health-bioscience"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="machine-intelligence"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="machine-perception"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="machine-translation"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="natural-language-processing"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="responsible-ai"> </li> <li class="glue-grid__col glue-grid__col--span-3 glue-grid__col--span-6-md glue-filter__result glue-is-hidden" data-glue-filter-="cloud-ai"> </li> </ul> </div> </div> </div> <div class="sort-wrapper"> <div class="mdc-select mdc-select--outlined glue-select sort-dropdown js-sort-dropdown" data-target-id="publications-index-list" > <input type="hidden" name="myselect" value=""> <div class="mdc-select__anchor not-glue" role="button" aria-haspopup="listbox" aria-labelledby="demo-label selectedtext-myselect" > <span id="demo-label" class="mdc-floating-label">Sort By</span> <span id="selectedtext-myselect" class="mdc-select__selected-text glue-grid__col--align-middle"></span> <span class="mdc-select__dropdown-icon"> <svg class="mdc-select__dropdown-icon-graphic" viewBox="7 10 10 5"> <polygon class="mdc-select__dropdown-icon-inactive" stroke="none" fill-rule="evenodd" points="7 10 12 15 17 10"></polygon> <polygon class="mdc-select__dropdown-icon-active" stroke="none" fill-rule="evenodd" points="7 15 12 10 17 15"></polygon> </svg> </span> </div> <div class="mdc-select__menu mdc-menu mdc-menu-surface"> <ul class="mdc-list" role="listbox"> <li class="mdc-list-item" aria-selected="false" data-value="title" role="option" aria-setsize="2" aria-posinset=1> <span class="mdc-list-item__ripple not-glue"></span> <span class="mdc-list-item__text">Title</span> </li> <li class="mdc-list-item" aria-selected="false" data-value="-title" role="option" aria-setsize="2" aria-posinset=2> <span class="mdc-list-item__ripple not-glue"></span> <span class="mdc-list-item__text">Title, descending</span> </li> <li class="mdc-list-item" aria-selected="false" data-value="publication__year" role="option" aria-setsize="2" aria-posinset=3> <span class="mdc-list-item__ripple not-glue"></span> <span class="mdc-list-item__text">Year</span> </li> <li class="mdc-list-item" aria-selected="false" data-value="-publication__year" role="option" aria-setsize="2" aria-posinset=4> <span class="mdc-list-item__ripple not-glue"></span> <span class="mdc-list-item__text">Year, descending</span> </li> </ul> </div> </div> </div> <div class="list-wrapper"> <div class="chips-wrapper spacer-03-bottom"> <div class="chipset" id="publications-index-chipset"> <div aria-hidden="true" class="chip-template"> <div class="chip" data-chip-value="template" data-chip-name=""> <div class="chip__label caption">chip template</div> <button class="chip__close" aria-label="Remove chip template"> <div class="glue-visually-hidden">Remove </div> <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--18px "> <use href="/gr/static/assets/icons/glue-icons.svg#close"></use> </svg> </button> </div> </div> <ul class="chipset-list" title="Applied filters"></ul> </div> </div> <div class="publications-list --theme- row-card-list" data-hot-swap="google-publications-list" data-gt-id="publications_list" data-gt-component-name=""> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/model-swarms-collaborative-search-of-adapted-llm-experts-via-swarm-intelligence/ > Model Swarms: Collaborative Search of Adapted LLM Experts via Swarm Intelligence </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> Shangbin Feng </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Yike Wang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/saynaebrahimi/"> Sayna Ebrahimi </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/hamidpalangi/"> Hamid Palangi </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/lesly/"> Lesly Miculicich </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Ace Kulshrestha </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Nathalie Rauschmayr </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Yejin Choi </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Yulia Tsvetkov </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentwe-propose-model-swarms-a-collabo tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentwe-propose-model-swarms-a-collabo" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">We propose Model Swarms, a collaborative search algorithm to adapt LLM experts via swarm intelligence. Specifically, Model Swarms starts with a pool of LLM experts and a utility function. Guided by the best-found checkpoints across models, diverse LLM experts collaboratively move in the weight space and search for adapted models that optimize the utility function. Compared to existing model composition approaches, Model Swarms offers modularity, works in low-data regimes, and doesn't need assumptions about existing experts and how they should be composed. Extensive experiments demonstrate that Model Swarms could flexibly adapt LLM experts to a single dataset, multi-dataset domains, reward models, as well as diverse human preferences. Further analysis reveals that LLM experts discover previously unseen capabilities in the search process and that Model Swarms enable the weak-to-strong transition of experts through the collaborative search process. </span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/model-swarms-collaborative-search-of-adapted-llm-experts-via-swarm-intelligence/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/aspest-bridging-the-gap-between-active-learning-and-selective-prediction/ > ASPEST: Bridging the Gap Between Active Learning and Selective Prediction </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/jiefengchen/"> Jiefeng Chen </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106672/"> Jinsung Yoon </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/saynaebrahimi/"> Sayna Ebrahimi </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106303/"> Sercan Arik </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Somesh Jha </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Transactions on Machine Learning Research (TMLR) (2024) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentselective-prediction-aims-to-learn tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentselective-prediction-aims-to-learn" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Selective prediction aims to learn a reliable model that abstains from making predictions when uncertain. These predictions can then be deferred to humans for further evaluation. As an everlasting challenge for machine learning, in many real-world scenarios, the distribution of test data is different from the training data. This results in more inaccurate predictions, and often increased dependence on humans, which can be difficult and expensive. Active learning aims to lower the overall labeling effort, and hence human dependence, by querying the most informative examples. Selective prediction and active learning have been approached from different angles, with the connection between them missing. In this work, we introduce a new learning paradigm, active selective prediction, which aims to query more informative samples from the shifted target domain while increasing accuracy and coverage. For this new paradigm, we propose a simple yet effective approach, ASPEST, that utilizes ensembles of model snapshots with self-training with their aggregated outputs as pseudo labels. Extensive experiments on numerous image, text and structured datasets, which suffer from domain shifts, demonstrate that ASPEST can significantly outperform prior work on selective prediction and active learning (e.g. on the MNIST→SVHN benchmark with the labeling budget of 100, ASPEST improves the AUACC metric from 79.36% to 88.84%) and achieves more optimal utilization of humans in the loop.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/aspest-bridging-the-gap-between-active-learning-and-selective-prediction/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/calm-contrasting-large-and-small-language-models-to-verify-grounded-generation/ > CaLM: Contrasting Large and Small Language Models to Verify Grounded Generation </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> I-Hung Hsu </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107150/"> Long Le </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/lesly/"> Lesly Miculicich </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Nanyun Peng </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentgrounded-generation-aims-to-equip tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentgrounded-generation-aims-to-equip" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Grounded generation aims to equip language models (LMs) with the ability to produce more credible and accountable responses by accurately citing verifiable sources. However, existing methods, by either feeding LMs with raw or preprocessed materials, remain prone to errors. To address this, we introduce CaLM, a novel verification framework. CaLM leverages the insight that a robust grounded response should be consistent with information derived solely from its cited sources. Our framework empowers smaller LMs, which rely less on parametric memory and excel at processing relevant information given a query, to validate the output of larger LMs. Larger LM responses that closely align with the smaller LMs' output, which relies exclusively on cited documents, are verified. Responses showing discrepancies are iteratively refined through a feedback loop. Experiments on three open-domain question-answering datasets demonstrate significant performance gains of 1.5% to 7% absolute average without any required model fine-tuning.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/calm-contrasting-large-and-small-language-models-to-verify-grounded-generation/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/found-in-the-middle-calibrating-positional-attention-bias-improves-long-context-utilization/ > Found in the middle: Calibrating Positional Attention Bias Improves Long Context Utilization </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> Cheng-Yu Hsieh </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Yung-Sung Chuang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Chun-Liang Li </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107150/"> Long Le </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Abhishek Kumar </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> James Glass </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Alexander Ratner </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Ranjay Krishna </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentlarge-language-models-llms-even tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentlarge-language-models-llms-even" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Large language models (LLMs), even when specifically trained to process long input contexts, struggle to capture relevant information located in the middle of their input. This phenomenon has been known as the lost-in-the-middle problem. In this work, we make three contributions. First, we set out to understand the factors that cause this phenomenon. In doing so, we establish a connection between lost-in-the-middle to LLMs' intrinsic attention bias: LLMs exhibit a U-shaped attention bias where the tokens at the beginning and at the end of its input receive higher attention, regardless of their relevance. Second, we mitigate this positional bias through a calibration mechanism, found-in-the-middle, that allows the model to attend to contexts faithfully according to their relevance, even though when they are in the middle. Third, we show found-in-the-middle not only achieves better performance in locating relevant information within a long context, but also eventually leads to improved retrieval-augmented generation (RAG) performance across various tasks, outperforming existing methods by up to 15 percentage points. These findings open up future directions in understanding LLM attention bias and its potential consequences.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/found-in-the-middle-calibrating-positional-attention-bias-improves-long-context-utilization/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/chain-of-table-evolves-tables-in-the-llm-reasoning-chain-for-table-understanding/ > Chain-of-Table: Evolves Tables in the LLM Reasoning Chain for Table Understanding </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> Zilong Wang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Hao Zhang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Chun-Liang Li </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106772/"> Julian Eisenschlos </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/vincentperot/"> Vincent Perot </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/lesly/"> Lesly Miculicich </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/yasuhisafujii/"> Yasuhisa Fujii </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Jingbo Shang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> ICLR (2024) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contenttable-based-reasoning-with-large-l tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contenttable-based-reasoning-with-large-l" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Table-based reasoning with large language models (LLMs) is a promising direction to tackle many table understanding tasks, such as table-based question answering and fact verification. Compared with generic reasoning, table-based reasoning requires the extraction of underlying semantics from both free-form questions and semi-structured tabular data. Chain-of-Thought and its similar approaches incorporate the reasoning chain in the form of textual context, but it is still an open question how to effectively leverage tabular data in the reasoning chain. We propose the Chain-of-Table framework, where tabular data is explicitly used in the reasoning chain as a proxy for intermediate thoughts. Specifically, we guide LLMs using in-context learning to iteratively generate operations and update the table to represent a tabular reasoning chain. LLMs can therefore dynamically plan the next operation based on the results of the previous ones. This continuous evolution of the table forms a chain, showing the reasoning process for a given tabular problem. The chain carries structured information of the intermediate results, enabling more accurate and reliable predictions. Chain-of-Table achieves new state-of-the-art performance on WikiTQ, FeTaQA, and TabFact benchmarks across multiple LLM choices.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/chain-of-table-evolves-tables-in-the-llm-reasoning-chain-for-table-understanding/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/codeclm-aligning-language-models-with-tailored-synthetic-data/ > CodecLM: Aligning Language Models with Tailored Synthetic Data </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Chun-Liang Li </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/vincentperot/"> Vincent Perot </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107150/"> Long Le </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Jin Miao </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106816/"> Zizhao Zhang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> NAACL 2024 </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentinstruction-tuning-has-emerged-as tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentinstruction-tuning-has-emerged-as" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Instruction tuning has emerged as the key in aligning large language models (LLMs) with specific task instructions, thereby mitigating the discrepancy between the next-token prediction objective and users' actual goals. To reduce the labor and time cost to collect or annotate data by humans, researchers start to explore the use of LLMs to generate instruction-aligned synthetic data. Recent works focus on generating diverse instructions and applying LLM to increase instruction complexity, often neglecting downstream use cases. It remains unclear how to tailor high-quality data to elicit better instruction-following abilities in different target instruction distributions and LLMs. To this end, we introduce CodecLM, a general framework for adaptively generating high-quality synthetic data for LLM alignment with different downstream instruction distributions and LLMs. Drawing on the Encode-Decode principles, we use LLMs as codecs to guide the data generation process. We first encode seed instructions into metadata, which are concise keywords generated on-the-fly to capture the target instruction distribution, and then decode metadata to create tailored instructions. We also introduce Self-Rubrics and Contrastive Filtering during decoding to tailor data-efficient samples. Extensive experiments on four open-domain instruction following benchmarks validate the effectiveness of CodecLM over the current state-of-the-arts.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/codeclm-aligning-language-models-with-tailored-synthetic-data/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/neural-spline-search-for-quantile-probabilistic-forecasting/ > Neural Spline Search for Quantile Probabilistic Forecasting </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107194/"> Ruoxi Sun </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Chun-Liang Li </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106303/"> Sercan Arik </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/mikedusenberry/"> Mike Dusenberry </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> AAAI (2023) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentaccurate-estimation-of-output-quan tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentaccurate-estimation-of-output-quan" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Accurate estimation of output quantiles is crucial in many use cases, where it is desired to model the range of possibility. Modeling target distribution at arbitrary quantile levels and at arbitrary input attribute levels are important to offer a comprehensive picture of the data, and requires the quantile function to be expressive enough. The quantile function describing the target distribution using quantile levels is critical for quantile regression. Althought various parametric forms for the distributions (that the quantile function specifies) can be adopted, an everlasting problem is selecting the most appropriate one that can properly approximate the data distributions. In this paper, we propose a non-parametric and data-driven approach, Neural Spline Search (NSS), to represent the observed data distribution without parametric assumptions. NSS is flexible and expressive for modeling data distributions by transforming the inputs with a series of monotonic spline regressions guided by symbolic operators. We demonstrate that NSS outperforms previous methods on synthetic, real-world regression and time-series forecasting tasks.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/neural-spline-search-for-quantile-probabilistic-forecasting/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/sqlprompt-improved-in-context-learning-for-few-shot-text-to-sql/ > SQLPrompt: Improved In-context Learning for Few-shot Text-to-SQL </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107194/"> Ruoxi Sun </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106303/"> Sercan Arik </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/rajarishisinha/"> Raj Sinha </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107950/"> Hootan Nakhost </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/hanjundai/"> Hanjun Dai </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/pengchengyin/"> Pengcheng Yin </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Findings of Conference on Empirical Methods in Natural Language Processing (EMNLP) (2023) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contenttext-to-sql-aims-to-automate-the-p tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contenttext-to-sql-aims-to-automate-the-p" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Text-to-SQL aims to automate the process of generating SQL queries on a database from natural language text. In this work, we propose "SQLPrompt", tailored to improve the few-shot prompting capabilities of Text-to-SQL for Large Language Models (LLMs). Our methods include innovative prompt design, execution based consistency decoding strategy which selects the SQL with the most consistent execution outcome among other SQL proposals, and a method that aims to improve performance by diversifying the SQL proposals during consistency selection with different prompt designs ("MixPrompt") and foundation models ("MixLLMs"). We show that SQLPrompt outperforms previous approaches for in-context learning with few labeled data by a large margin, closing the gap with finetuning state-of the-art with thousands of labeled data.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/sqlprompt-improved-in-context-learning-for-few-shot-text-to-sql/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/universal-self-adaptive-prompting/ > Universal Self-adaptive Prompting </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/xingchenwan/"> Xingchen Wan </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107194/"> Ruoxi Sun </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107950/"> Hootan Nakhost </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/hanjundai/"> Hanjun Dai </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106772/"> Julian Eisenschlos </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106303/"> Sercan Arik </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Empirical Methods in Natural Language Processing (EMNLP) (2023) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contenta-hallmark-of-modern-large-languag tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contenta-hallmark-of-modern-large-languag" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">A hallmark of modern large language models (LLMs) is their impressive general zero-shot and few-shot abilities, often elicited through in-context learning (ICL) via prompting. However, while highly coveted and being the most general, zero-shot performances in LLMs are still typically weaker due to the lack of guidance and the difficulty of applying existing automatic prompt design methods in general tasks when ground-truth labels are unavailable. In this study, we address this by presenting Universal Self-Adaptive Prompting (USP), an automatic prompt design approach specifically tailored for zero-shot learning (while compatible with few-shot). Requiring only a small amount of unlabeled data and an inference-only LLM, USP is highly versatile: to achieve universal prompting, USP categorizes a possible NLP task into one of the three possible task types and then uses a corresponding selector to select the most suitable queries and zero-shot model-generated responses as pseudo-demonstrations, thereby generalizing ICL to the zero-shot setup in a fully automated way. We evaluate USP with PaLM and PaLM 2 models and demonstrate performances that are considerably stronger than standard zero-shot baselines and often comparable to or even superior to few-shot baselines across more than 40 natural language understanding, natural language generation, and reasoning tasks.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/universal-self-adaptive-prompting/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> <div class="row-card"> <div class="row-card__container"> <div class="row-card__body"> <a class="row-card__heading headline-6 glue-link" href=http://research.google/pubs/queryform-a-simple-zero-shot-form-entity-query-framework/ > QueryForm: A Simple Zero-shot Form Entity Query Framework </a> <div class="row-card__subheading"> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/108469/"> Zifeng Wang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/106816/"> Zizhao Zhang </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Jacob Devlin </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/107149/"> Chen-Yu Lee </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/guolongsu/"> Guolong Su </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Hao Zhang </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> Jennifer Dy </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/vincentperot/"> Vincent Perot </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> <a class="row-card__small-link" href="/people/105803/"> Tomas Pfister </a> </div> <div class="row-card__subheading__spacer"></div> <div class="row-card__subheading__item extra-small-text"> ACL (2023) </div> </div> </div> <div class="row-card__cta headline-6"> <div class="glue-tooltip" data-glue-tooltip-auto-position="false"> <button class="glue-button glue-button--low-emphasis glue-tooltip__trigger" aria-describedby=tooltip-contentzero-shot-transfer-learning-for-do tabindex=0 > <span class="js-gt-item-id">Preview</span> </button> <span id="tooltip-contentzero-shot-transfer-learning-for-do" class="glue-tooltip__content" role="tooltip"> <span data-tooltip-type="simple"> Preview abstract </span> <span data-tooltip-type="rich"> <span class="glue-tooltip__body">Zero-shot transfer learning for document understanding is a crucial yet under-investigated scenario to help reduce the high cost involved in annotating document entities. We present a novel query-based framework, QueryForm, that extracts entity values from form-like documents in a zero-shot fashion. QueryForm contains a dual prompting mechanism that composes both the document schema and a specific entity type into a query, which is used to prompt a Transformer model to perform a single entity extraction task. Furthermore, we propose to leverage large-scale query-entity pairs generated from form-like webpages with weak HTML annotations to pre-train QueryForm. By unifying pre-training and fine-tuning into the same query-based framework, QueryForm enables models to learn from structured documents containing various entities and layouts, leading to better generalization to target document types without the need for target-specific training data. QueryForm sets new state-of-the-art average F1 score on both the XFUND (+4.6%~10.1%) and the Payment (+3.2%~9.5%) zero-shot benchmark, with a smaller model size and no additional image input.</span> <a class="glue-button glue-button--low-emphasis" href="http://research.google/pubs/queryform-a-simple-zero-shot-form-entity-query-framework/" > <span class="js-gt-item-id">View details</span> </a> </span> </span> </div> </div> </div> </div> </div> <div class="filtered-list-base__pagination" data-hot-swap="pagination"> <nav class="pagination" aria-label="pagination" data-gt-id="Tomas Pfister_pagination" data-gt-component-name="TODO: add BE var"> <a href="javascript(0):void" rel="nofollow" title="Go to page 1" aria-label="Go to page 1, first page" data-page="1" class="pagination__first-button js-pagination-button pagination__first-button--disabled" aria-disabled="true" > <svg role="presentation" aria-hidden="true" xmlns="http://www.w3.org/2000/svg" height="24px" viewBox="0 0 24 24" width="24px" fill="#000000"> <path d="M18.41 16.59L13.82 12l4.59-4.59L17 6l-6 6 6 6zM6 6h2v12H6z" /> <path d="M24 24H0V0h24v24z" fill="none" /> </svg> </a> <a href="javascript(0):void" rel="nofollow" title="Go to page 1" aria-label="Go to page 1" data-page="0" class="pagination__back-button js-pagination-button pagination__back-button--disabled " aria-disabled="true" > <svg role="presentation" aria-hidden="true" aria-hidden="true" class="glue-icon glue-icon--24px "> <use href="/gr/static/assets/icons/glue-icons.svg#chevron-left"></use> </svg> </a> <ul class="pagination__list"> <li class="pagination__item pagination__item--current"> <a href="javascript(0):void" rel="nofollow" title="Go to page 1, current page" aria-current="page 1, current page" data-page="1" class="pagination__link js-pagination-button">1</a> </li> <li class="pagination__item"> <a href="javascript(0):void" rel="nofollow" aria-label="Go to page 2" title="Go to page 2" data-page="2" class="pagination__link js-pagination-button">2</a> </li> <li class="pagination__item"> <a href="javascript(0):void" rel="nofollow" aria-label="Go to page 3" title="Go to page 3" data-page="3" class="pagination__link js-pagination-button">3</a> </li> <li class="pagination__ellipsis">…</li> </ul> <form class="pagination__form-mobile js-pagination-form-mobile" data-max-pages="6"> <label for="page">of 6</label> <input id="page-mobile" type="number" min="0" name="page" value="1" class="js-pagination-input-mobile"></input> </form> <a href="javascript(0):void" rel="nofollow" aria-label="Go to page 2" title="Go to page 2" data-page="2" data-max-pages="6" class="pagination__next-button js-pagination-button " > <svg role="presentation" aria-hidden="true" aria-hidden="true" class="glue-icon glue-icon--24px "> <use href="/gr/static/assets/icons/glue-icons.svg#chevron-right"></use> </svg> </a> <a href="javascript(0):void" rel="nofollow" aria-label="Go to page 6, last page" title="Go to page 6, last page" data-page="6" class="pagination__last-button js-pagination-button " > <svg role="presentation" aria-hidden="true" xmlns="http://www.w3.org/2000/svg" height="24px" viewBox="0 0 24 24" width="24px" fill="#000000"> <path d="M0 0h24v24H0V0z" fill="none" /> <path d="M5.59 7.41L10.18 12l-4.59 4.59L7 18l6-6-6-6zM16 6h2v12h-2z" /> </svg> </a> <form class="pagination__form js-pagination-form" data-max-pages="6"> <input id="page-desktop" aria-label="Page number" type="number" min="0" name="page" class="js-pagination-input" value="1"></input> <label>of 6 pages</label> </form> </nav> </div> </div> </div> </section> <section class="google_scholar_search">  <a href="#" class="js-google-scholar-search-link glue-button glue-button--low-emphasis glue-button--icon" title="See more related publications on Google Scholar" target="_blank"> <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--24px "> <use href="/gr/static/assets/icons/glue-icons.svg#open-in-new"></use> </svg> Search on Google Scholar </a> </section> </div> <section class="banner --theme-" data-gt-id="banner" data-gt-component-name=""> <div class="banner__wrapper glue-page glue-grid"> <div class="banner__copy glue-grid__col glue-grid__col--span-4-sm glue-grid__col--span-6-md"> <h2 class="banner__headline headline-3">Join us</h2> <p class="banner__body-copy body">We're always looking for more talented, passionate people.</p> <a class="glue-button glue-button--medium-emphasis" href="http://research.google/careers/" > <span class="js-gt-item-id">See opportunities</span> </a> </div> <div class="glue-grid__col glue-grid__col--span-0-sm glue-grid__col--span-1"></div> <div class="banner__image glue-grid__col glue-grid__col--span-4-sm glue-grid__col--span-5-md"> <img src="https://storage.googleapis.com/gweb-research2023-media/images/Careers.original.jpg" alt="Careers" /> </div> </div> </section> </main> <footer class="glue-footer"> <div class="glue-page"> <section class="glue-social"> <div class="glue-social__group glue-social--monochrome"> <p class="glue-social__title glue-social__title--inline"> Follow us </p> <nav class="js-gt-follow-us-wrapper" aria-label="Social media links"> <ul class="glue-social__list" role="list"> <li class="glue-social__item"> <a class="glue-social__link" href="https://twitter.com/GoogleAI" title="Follow us on x" target="_blank" rel="noopener" data-gt-method="x"" > <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--social glue-icon--24px"> <use href="/gr/static/assets/icons/twitter-x.svg#twitter-x"></use> </svg> </a> </li> <li class="glue-social__item"> <a class="glue-social__link" href="https://www.linkedin.com/showcase/googleresearch/" title="Follow us on linkedin" target="_blank" rel="noopener" data-gt-method="linkedin"" > <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--social glue-icon--24px"> <use href="/gr/static/assets/icons/glue-icons.svg#post-linkedin"></use> </svg> </a> </li> <li class="glue-social__item"> <a class="glue-social__link" href="https://www.youtube.com/c/GoogleResearch" title="Follow us on youtube" target="_blank" rel="noopener" data-gt-method="youtube"" > <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--social glue-icon--24px"> <use href="/gr/static/assets/icons/glue-icons.svg#video-youtube"></use> </svg> </a> </li> <li class="glue-social__item"> <a class="glue-social__link" href="https://github.com/google-research" title="Follow us on github" target="_blank" rel="noopener" data-gt-method="github"" > <svg role="presentation" aria-hidden="true" class="glue-icon glue-icon--social glue-icon--24px"> <use href="/gr/static/assets/icons/github.svg#github"></use> </svg> </a> </li> </ul> </nav> </div> </section> </div> <div class="glue-fullbleed"></div> <section class="glue-page"> <nav class="glue-footer__global" aria-label="Footer resource links"> <div class="glue-footer__logo"> <a href="https://www.google.com" title="Google" class="glue-footer__link"> <svg role="presentation" aria-hidden="true" class="glue-icon glue-footer__logo-img"> <use href="/gr/static/assets/icons/glue-icons.svg#google-solid-logo"></use> </svg> </a> </div> <ul class="glue-footer__global-links glue-no-bullet js-gt-global-nav-wrapper" role="list"> <li class="glue-footer__global-links-list-item" data-gt-primary="About Google"> <a class="glue-footer__link" href="https://about.google/" target="_blank" rel="noopener"> About Google </a> </li> <li class="glue-footer__global-links-list-item" data-gt-primary="Google Products"> <a class="glue-footer__link" href="https://about.google/intl/en/products/" target="_blank" rel="noopener"> Google Products </a> </li> <li class="glue-footer__global-links-list-item" data-gt-primary="Privacy"> <a class="glue-footer__link" href="https://policies.google.com/privacy" target="_blank" rel="noopener"> Privacy </a> </li> <li class="glue-footer__global-links-list-item" data-gt-primary="Terms"> <a class="glue-footer__link" href="https://policies.google.com/terms" target="_blank" rel="noopener"> Terms </a> </li> </ul> <ul class="glue-footer__global-links glue-footer__global-links--extra glue-no-bullet" role="list"> <li class="glue-footer__global-links-list-item glue-footer__global-links-list-item--extra"> <a class="glue-footer__link" href="https://support.google.com/?hl=en"> <svg role="presentation" aria-hidden="true" aria-hidden="true" class="glue-icon glue-icon--24px glue-icon--footer-help"> <use href="/gr/static/assets/icons/glue-icons.svg#help"></use> </svg> Help </a> </li> <li class="glue-footer__global-links-list-item glue-footer__global-links-list-item--extra"> <button class="glue-footer__link google-feedback js-feedback-button" href="" data-product-id="5137383" > Submit feedback </button> </li> </ul> </nav> </section> </footer> <script src="https://www.gstatic.com/glue/v27_1/material-components-web.min.js"></script> <script src="https://www.youtube.com/player_api"></script> <script type="text/javascript" src="/gr/static/js/googleresearch.js?id=b70549917812130af912601ad763f13e"></script> <script type="text/javascript" src="https://support.google.com/inapp/api.js"></script> <script src="https://www.gstatic.com/glue/cookienotificationbar/cookienotificationbar.min.js" data-glue-cookie-notification-bar-category="2B"> </script> </body> </html>

CINXE.COM

Tomas Pfister