Search | arXiv e-print repository

<!DOCTYPE html> <html lang="en"> <head> <meta charset="utf-8"/> <meta name="viewport" content="width=device-width, initial-scale=1"/>  <link rel="apple-touch-icon" sizes="180x180" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/apple-touch-icon.png"> <link rel="icon" type="image/png" sizes="32x32" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon-32x32.png"> <link rel="icon" type="image/png" sizes="16x16" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon-16x16.png"> <link rel="manifest" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/site.webmanifest"> <link rel="mask-icon" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/safari-pinned-tab.svg" color="#b31b1b"> <link rel="shortcut icon" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon.ico"> <meta name="msapplication-TileColor" content="#b31b1b"> <meta name="msapplication-config" content="images/icons/browserconfig.xml"> <meta name="theme-color" content="#b31b1b">  <title>Search | arXiv e-print repository</title> <script defer src="https://static.arxiv.org/static/base/1.0.0a5/fontawesome-free-5.11.2-web/js/all.js"></script> <link rel="stylesheet" href="https://static.arxiv.org/static/base/1.0.0a5/css/arxivstyle.css" /> <script type="text/x-mathjax-config"> MathJax.Hub.Config({ messageStyle: "none", extensions: ["tex2jax.js"], jax: ["input/TeX", "output/HTML-CSS"], tex2jax: { inlineMath: [ ['$','$'], ["\$","\$"] ], displayMath: [ ['$$','$$'], ["\\[","\\]"] ], processEscapes: true, ignoreClass: '.*', processClass: 'mathjax.*' }, TeX: { extensions: ["AMSmath.js", "AMSsymbols.js", "noErrors.js"], noErrors: { inlineDelimiters: ["$","$"], multiLine: false, style: { "font-size": "normal", "border": "" } } }, "HTML-CSS": { availableFonts: ["TeX"] } }); </script> <script src='//static.arxiv.org/MathJax-2.7.3/MathJax.js'></script> <script src="https://static.arxiv.org/static/base/1.0.0a5/js/notification.js"></script> <link rel="stylesheet" href="https://static.arxiv.org/static/search/0.5.6/css/bulma-tooltip.min.css" /> <link rel="stylesheet" href="https://static.arxiv.org/static/search/0.5.6/css/search.css" /> <script src="https://code.jquery.com/jquery-3.2.1.slim.min.js" integrity="sha256-k2WSCIexGzOj3Euiig+TlR8gA0EmPjuc79OEeY5L45g=" crossorigin="anonymous"></script> <script src="https://static.arxiv.org/static/search/0.5.6/js/fieldset.js"></script> <style> radio#cf-customfield_11400 { display: none; } </style> </head> <body> <header><a href="#main-container" class="is-sr-only">Skip to main content</a>  <div class="attribution level is-marginless" role="banner"> <div class="level-left"> <a class="level-item" href="https://cornell.edu/"><img src="https://static.arxiv.org/static/base/1.0.0a5/images/cornell-reduced-white-SMALL.svg" alt="Cornell University" width="200" aria-label="logo" /></a> </div> <div class="level-right is-marginless"><p class="sponsors level-item is-marginless"><span id="support-ack-url">We gratefully acknowledge support from<br /> the Simons Foundation, <a href="https://info.arxiv.org/about/ourmembers.html">member institutions</a>, and all contributors. <a href="https://info.arxiv.org/about/donate.html">Donate</a></span></p></div> </div>  <div class="identity level is-marginless"> <div class="level-left"> <div class="level-item"> <a class="arxiv" href="https://arxiv.org/" aria-label="arxiv-logo"> <img src="https://static.arxiv.org/static/base/1.0.0a5/images/arxiv-logo-one-color-white.svg" aria-label="logo" alt="arxiv logo" width="85" style="width:85px;"/> </a> </div> </div> <div class="search-block level-right"> <form class="level-item mini-search" method="GET" action="https://arxiv.org/search"> <div class="field has-addons"> <div class="control"> <input class="input is-small" type="text" name="query" placeholder="Search..." aria-label="Search term or terms" /> <p class="help"><a href="https://info.arxiv.org/help">Help</a> | <a href="https://arxiv.org/search/advanced">Advanced Search</a></p> </div> <div class="control"> <div class="select is-small"> <select name="searchtype" aria-label="Field to search"> <option value="all" selected="selected">All fields</option> <option value="title">Title</option> <option value="author">Author</option> <option value="abstract">Abstract</option> <option value="comments">Comments</option> <option value="journal_ref">Journal reference</option> <option value="acm_class">ACM classification</option> <option value="msc_class">MSC classification</option> <option value="report_num">Report number</option> <option value="paper_id">arXiv identifier</option> <option value="doi">DOI</option> <option value="orcid">ORCID</option> <option value="author_id">arXiv author ID</option> <option value="help">Help pages</option> <option value="full_text">Full text</option> </select> </div> </div> <input type="hidden" name="source" value="header"> <button class="button is-small is-cul-darker">Search</button> </div> </form> </div> </div>  <div class="container"> <div class="user-tools is-size-7 has-text-right has-text-weight-bold" role="navigation" aria-label="User menu"> <a href="https://arxiv.org/login">Login</a> </div> </div> </header> <main class="container" id="main-container"> <div class="level is-marginless"> <div class="level-left"> <h1 class="title is-clearfix"> Showing 1–24 of 24 results for author: <span class="mathjax">Shoham, Y</span> </h1> </div> <div class="level-right is-hidden-mobile">  <span class="help" style="display: inline-block;"><a href="https://github.com/arXiv/arxiv-search/releases">Search v0.5.6 released 2020-02-24</a>  </span> </div> </div> <div class="content"> <form method="GET" action="/search/cs" aria-role="search"> Searching in archive <strong>cs</strong>. <a href="/search/?searchtype=author&query=Shoham%2C+Y">Search in all archives.</a> <div class="field has-addons-tablet"> <div class="control is-expanded"> <label for="query" class="hidden-label">Search term or terms</label> <input class="input is-medium" id="query" name="query" placeholder="Search term..." type="text" value="Shoham, Y"> </div> <div class="select control is-medium"> <label class="is-hidden" for="searchtype">Field</label> <select class="is-medium" id="searchtype" name="searchtype"><option value="all">All fields</option><option value="title">Title</option><option selected value="author">Author(s)</option><option value="abstract">Abstract</option><option value="comments">Comments</option><option value="journal_ref">Journal reference</option><option value="acm_class">ACM classification</option><option value="msc_class">MSC classification</option><option value="report_num">Report number</option><option value="paper_id">arXiv identifier</option><option value="doi">DOI</option><option value="orcid">ORCID</option><option value="license">License (URI)</option><option value="author_id">arXiv author ID</option><option value="help">Help pages</option><option value="full_text">Full text</option></select> </div> <div class="control"> <button class="button is-link is-medium">Search</button> </div> </div> <div class="field"> <div class="control is-size-7"> <label class="radio"> <input checked id="abstracts-0" name="abstracts" type="radio" value="show"> Show abstracts </label> <label class="radio"> <input id="abstracts-1" name="abstracts" type="radio" value="hide"> Hide abstracts </label> </div> </div> <div class="is-clearfix" style="height: 2.5em"> <div class="is-pulled-right"> <a href="/search/advanced?terms-0-term=Shoham%2C+Y&terms-0-field=author&size=50&order=-announced_date_first">Advanced Search</a> </div> </div> <input type="hidden" name="order" value="-announced_date_first"> <input type="hidden" name="size" value="50"> </form> <div class="level breathe-horizontal"> <div class="level-left"> <form method="GET" action="/search/"> <div style="display: none;"> <select id="searchtype" name="searchtype"><option value="all">All fields</option><option value="title">Title</option><option selected value="author">Author(s)</option><option value="abstract">Abstract</option><option value="comments">Comments</option><option value="journal_ref">Journal reference</option><option value="acm_class">ACM classification</option><option value="msc_class">MSC classification</option><option value="report_num">Report number</option><option value="paper_id">arXiv identifier</option><option value="doi">DOI</option><option value="orcid">ORCID</option><option value="license">License (URI)</option><option value="author_id">arXiv author ID</option><option value="help">Help pages</option><option value="full_text">Full text</option></select> <input id="query" name="query" type="text" value="Shoham, Y"> <ul id="abstracts"><li><input checked id="abstracts-0" name="abstracts" type="radio" value="show"> <label for="abstracts-0">Show abstracts</label></li><li><input id="abstracts-1" name="abstracts" type="radio" value="hide"> <label for="abstracts-1">Hide abstracts</label></li></ul> </div> <div class="box field is-grouped is-grouped-multiline level-item"> <div class="control"> <span class="select is-small"> <select id="size" name="size"><option value="25">25</option><option selected value="50">50</option><option value="100">100</option><option value="200">200</option></select> </span> <label for="size">results per page</label>. </div> <div class="control"> <label for="order">Sort results by</label> <span class="select is-small"> <select id="order" name="order"><option selected value="-announced_date_first">Announcement date (newest first)</option><option value="announced_date_first">Announcement date (oldest first)</option><option value="-submitted_date">Submission date (newest first)</option><option value="submitted_date">Submission date (oldest first)</option><option value="">Relevance</option></select> </span> </div> <div class="control"> <button class="button is-small is-link">Go</button> </div> </div> </form> </div> </div> <ol class="breathe-horizontal" start="1"> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.12570">arXiv:2408.12570</a> <span> [<a href="https://arxiv.org/pdf/2408.12570">pdf</a>, <a href="https://arxiv.org/format/2408.12570">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Jamba-1.5: Hybrid Transformer-Mamba Models at Scale </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jamba+Team"> Jamba Team</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Arazi%2C+A">Alan Arazi</a>, <a href="/search/cs?searchtype=author&query=Bergman%2C+A">Amir Bergman</a>, <a href="/search/cs?searchtype=author&query=Manevich%2C+A">Avshalom Manevich</a>, <a href="/search/cs?searchtype=author&query=Peleg%2C+B">Barak Peleg</a>, <a href="/search/cs?searchtype=author&query=Aviram%2C+B">Ben Aviram</a>, <a href="/search/cs?searchtype=author&query=Almagor%2C+C">Chen Almagor</a>, <a href="/search/cs?searchtype=author&query=Fridman%2C+C">Clara Fridman</a>, <a href="/search/cs?searchtype=author&query=Padnos%2C+D">Dan Padnos</a>, <a href="/search/cs?searchtype=author&query=Gissin%2C+D">Daniel Gissin</a>, <a href="/search/cs?searchtype=author&query=Jannai%2C+D">Daniel Jannai</a>, <a href="/search/cs?searchtype=author&query=Muhlgay%2C+D">Dor Muhlgay</a>, <a href="/search/cs?searchtype=author&query=Zimberg%2C+D">Dor Zimberg</a>, <a href="/search/cs?searchtype=author&query=Gerber%2C+E+M">Edden M Gerber</a>, <a href="/search/cs?searchtype=author&query=Dolev%2C+E">Elad Dolev</a>, <a href="/search/cs?searchtype=author&query=Krakovsky%2C+E">Eran Krakovsky</a>, <a href="/search/cs?searchtype=author&query=Safahi%2C+E">Erez Safahi</a>, <a href="/search/cs?searchtype=author&query=Schwartz%2C+E">Erez Schwartz</a>, <a href="/search/cs?searchtype=author&query=Cohen%2C+G">Gal Cohen</a>, <a href="/search/cs?searchtype=author&query=Shachaf%2C+G">Gal Shachaf</a>, <a href="/search/cs?searchtype=author&query=Rozenblum%2C+H">Haim Rozenblum</a>, <a href="/search/cs?searchtype=author&query=Bata%2C+H">Hofit Bata</a>, <a href="/search/cs?searchtype=author&query=Blass%2C+I">Ido Blass</a>, <a href="/search/cs?searchtype=author&query=Magar%2C+I">Inbal Magar</a> , et al. (36 additional authors not shown) </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2408.12570v1-abstract-short" style="display: inline;"> We present Jamba-1.5, new instruction-tuned large language models based on our Jamba architecture. Jamba is a hybrid Transformer-Mamba mixture of experts architecture, providing high throughput and low memory usage across context lengths, while retaining the same or better quality as Transformer models. We release two model sizes: Jamba-1.5-Large, with 94B active parameters, and Jamba-1.5-Mini, wi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.12570v1-abstract-full').style.display = 'inline'; document.getElementById('2408.12570v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.12570v1-abstract-full" style="display: none;"> We present Jamba-1.5, new instruction-tuned large language models based on our Jamba architecture. Jamba is a hybrid Transformer-Mamba mixture of experts architecture, providing high throughput and low memory usage across context lengths, while retaining the same or better quality as Transformer models. We release two model sizes: Jamba-1.5-Large, with 94B active parameters, and Jamba-1.5-Mini, with 12B active parameters. Both models are fine-tuned for a variety of conversational and instruction-following capabilties, and have an effective context length of 256K tokens, the largest amongst open-weight models. To support cost-effective inference, we introduce ExpertsInt8, a novel quantization technique that allows fitting Jamba-1.5-Large on a machine with 8 80GB GPUs when processing 256K-token contexts without loss of quality. When evaluated on a battery of academic and chatbot benchmarks, Jamba-1.5 models achieve excellent results while providing high throughput and outperforming other open-weight models on long-context benchmarks. The model weights for both sizes are publicly available under the Jamba Open Model License and we release ExpertsInt8 as open source. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.12570v1-abstract-full').style.display = 'none'; document.getElementById('2408.12570v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 22 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Webpage: https://www.ai21.com/jamba</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.10937">arXiv:2406.10937</a> <span> [<a href="https://arxiv.org/pdf/2406.10937">pdf</a>, <a href="https://arxiv.org/ps/2406.10937">ps</a>, <a href="https://arxiv.org/format/2406.10937">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Understanding Understanding: A Pragmatic Framework Motivated by Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2406.10937v2-abstract-short" style="display: inline;"> Motivated by the rapid ascent of Large Language Models (LLMs) and debates about the extent to which they possess human-level qualities, we propose a framework for testing whether any agent (be it a machine or a human) understands a subject matter. In Turing-test fashion, the framework is based solely on the agent's performance, and specifically on how well it answers questions. Elements of the fra… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10937v2-abstract-full').style.display = 'inline'; document.getElementById('2406.10937v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.10937v2-abstract-full" style="display: none;"> Motivated by the rapid ascent of Large Language Models (LLMs) and debates about the extent to which they possess human-level qualities, we propose a framework for testing whether any agent (be it a machine or a human) understands a subject matter. In Turing-test fashion, the framework is based solely on the agent's performance, and specifically on how well it answers questions. Elements of the framework include circumscribing the set of questions (the "scope of understanding"), requiring general competence ("passing grade"), avoiding "ridiculous answers", but still allowing wrong and "I don't know" answers to some questions. Reaching certainty about these conditions requires exhaustive testing of the questions which is impossible for nontrivial scopes, but we show how high confidence can be achieved via random sampling and the application of probabilistic confidence bounds. We also show that accompanying answers with explanations can improve the sample complexity required to achieve acceptable bounds, because an explanation of an answer implies the ability to answer many similar questions. According to our framework, current LLMs cannot be said to understand nontrivial domains, but as the framework provides a practical recipe for testing understanding, it thus also constitutes a tool for building AI agents that do understand. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10937v2-abstract-full').style.display = 'none'; document.getElementById('2406.10937v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 19 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 16 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2405.19522">arXiv:2405.19522</a> <span> [<a href="https://arxiv.org/pdf/2405.19522">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Artificial Intelligence Index Report 2024 </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Maslej%2C+N">Nestor Maslej</a>, <a href="/search/cs?searchtype=author&query=Fattorini%2C+L">Loredana Fattorini</a>, <a href="/search/cs?searchtype=author&query=Perrault%2C+R">Raymond Perrault</a>, <a href="/search/cs?searchtype=author&query=Parli%2C+V">Vanessa Parli</a>, <a href="/search/cs?searchtype=author&query=Reuel%2C+A">Anka Reuel</a>, <a href="/search/cs?searchtype=author&query=Brynjolfsson%2C+E">Erik Brynjolfsson</a>, <a href="/search/cs?searchtype=author&query=Etchemendy%2C+J">John Etchemendy</a>, <a href="/search/cs?searchtype=author&query=Ligett%2C+K">Katrina Ligett</a>, <a href="/search/cs?searchtype=author&query=Lyons%2C+T">Terah Lyons</a>, <a href="/search/cs?searchtype=author&query=Manyika%2C+J">James Manyika</a>, <a href="/search/cs?searchtype=author&query=Niebles%2C+J+C">Juan Carlos Niebles</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Wald%2C+R">Russell Wald</a>, <a href="/search/cs?searchtype=author&query=Clark%2C+J">Jack Clark</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2405.19522v1-abstract-short" style="display: inline;"> The 2024 Index is our most comprehensive to date and arrives at an important moment when AI's influence on society has never been more pronounced. This year, we have broadened our scope to more extensively cover essential trends such as technical advancements in AI, public perceptions of the technology, and the geopolitical dynamics surrounding its development. Featuring more original data than ev… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2405.19522v1-abstract-full').style.display = 'inline'; document.getElementById('2405.19522v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2405.19522v1-abstract-full" style="display: none;"> The 2024 Index is our most comprehensive to date and arrives at an important moment when AI's influence on society has never been more pronounced. This year, we have broadened our scope to more extensively cover essential trends such as technical advancements in AI, public perceptions of the technology, and the geopolitical dynamics surrounding its development. Featuring more original data than ever before, this edition introduces new estimates on AI training costs, detailed analyses of the responsible AI landscape, and an entirely new chapter dedicated to AI's impact on science and medicine. The AI Index report tracks, collates, distills, and visualizes data related to artificial intelligence (AI). Our mission is to provide unbiased, rigorously vetted, broadly sourced data in order for policymakers, researchers, executives, journalists, and the general public to develop a more thorough and nuanced understanding of the complex field of AI. The AI Index is recognized globally as one of the most credible and authoritative sources for data and insights on artificial intelligence. Previous editions have been cited in major newspapers, including the The New York Times, Bloomberg, and The Guardian, have amassed hundreds of academic citations, and been referenced by high-level policymakers in the United States, the United Kingdom, and the European Union, among other places. This year's edition surpasses all previous ones in size, scale, and scope, reflecting the growing significance that AI is coming to hold in all of our lives. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2405.19522v1-abstract-full').style.display = 'none'; document.getElementById('2405.19522v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 29 May, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2403.19887">arXiv:2403.19887</a> <span> [<a href="https://arxiv.org/pdf/2403.19887">pdf</a>, <a href="https://arxiv.org/format/2403.19887">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Jamba: A Hybrid Transformer-Mamba Language Model </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lieber%2C+O">Opher Lieber</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Bata%2C+H">Hofit Bata</a>, <a href="/search/cs?searchtype=author&query=Cohen%2C+G">Gal Cohen</a>, <a href="/search/cs?searchtype=author&query=Osin%2C+J">Jhonathan Osin</a>, <a href="/search/cs?searchtype=author&query=Dalmedigos%2C+I">Itay Dalmedigos</a>, <a href="/search/cs?searchtype=author&query=Safahi%2C+E">Erez Safahi</a>, <a href="/search/cs?searchtype=author&query=Meirom%2C+S">Shaked Meirom</a>, <a href="/search/cs?searchtype=author&query=Belinkov%2C+Y">Yonatan Belinkov</a>, <a href="/search/cs?searchtype=author&query=Shalev-Shwartz%2C+S">Shai Shalev-Shwartz</a>, <a href="/search/cs?searchtype=author&query=Abend%2C+O">Omri Abend</a>, <a href="/search/cs?searchtype=author&query=Alon%2C+R">Raz Alon</a>, <a href="/search/cs?searchtype=author&query=Asida%2C+T">Tomer Asida</a>, <a href="/search/cs?searchtype=author&query=Bergman%2C+A">Amir Bergman</a>, <a href="/search/cs?searchtype=author&query=Glozman%2C+R">Roman Glozman</a>, <a href="/search/cs?searchtype=author&query=Gokhman%2C+M">Michael Gokhman</a>, <a href="/search/cs?searchtype=author&query=Manevich%2C+A">Avashalom Manevich</a>, <a href="/search/cs?searchtype=author&query=Ratner%2C+N">Nir Ratner</a>, <a href="/search/cs?searchtype=author&query=Rozen%2C+N">Noam Rozen</a>, <a href="/search/cs?searchtype=author&query=Shwartz%2C+E">Erez Shwartz</a>, <a href="/search/cs?searchtype=author&query=Zusman%2C+M">Mor Zusman</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2403.19887v2-abstract-short" style="display: inline;"> We present Jamba, a new base large language model based on a novel hybrid Transformer-Mamba mixture-of-experts (MoE) architecture. Specifically, Jamba interleaves blocks of Transformer and Mamba layers, enjoying the benefits of both model families. MoE is added in some of these layers to increase model capacity while keeping active parameter usage manageable. This flexible architecture allows reso… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2403.19887v2-abstract-full').style.display = 'inline'; document.getElementById('2403.19887v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2403.19887v2-abstract-full" style="display: none;"> We present Jamba, a new base large language model based on a novel hybrid Transformer-Mamba mixture-of-experts (MoE) architecture. Specifically, Jamba interleaves blocks of Transformer and Mamba layers, enjoying the benefits of both model families. MoE is added in some of these layers to increase model capacity while keeping active parameter usage manageable. This flexible architecture allows resource- and objective-specific configurations. In the particular configuration we have implemented, we end up with a powerful model that fits in a single 80GB GPU. Built at large scale, Jamba provides high throughput and small memory footprint compared to vanilla Transformers, and at the same time state-of-the-art performance on standard language model benchmarks and long-context evaluations. Remarkably, the model presents strong results for up to 256K tokens context length. We study various architectural decisions, such as how to combine Transformer and Mamba layers, and how to mix experts, and show that some of them are crucial in large scale modeling. We also describe several interesting properties of these architectures which the training and evaluation of Jamba have revealed, and plan to release checkpoints from various ablation runs, to encourage further exploration of this novel architecture. We make the weights of our implementation of Jamba publicly available under a permissive license. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2403.19887v2-abstract-full').style.display = 'none'; document.getElementById('2403.19887v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 3 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 28 March, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Webpage: https://www.ai21.com/jamba</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2310.03715">arXiv:2310.03715</a> <span> [<a href="https://arxiv.org/pdf/2310.03715">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> </div> </div> <p class="title is-5 mathjax"> Artificial Intelligence Index Report 2023 </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Maslej%2C+N">Nestor Maslej</a>, <a href="/search/cs?searchtype=author&query=Fattorini%2C+L">Loredana Fattorini</a>, <a href="/search/cs?searchtype=author&query=Brynjolfsson%2C+E">Erik Brynjolfsson</a>, <a href="/search/cs?searchtype=author&query=Etchemendy%2C+J">John Etchemendy</a>, <a href="/search/cs?searchtype=author&query=Ligett%2C+K">Katrina Ligett</a>, <a href="/search/cs?searchtype=author&query=Lyons%2C+T">Terah Lyons</a>, <a href="/search/cs?searchtype=author&query=Manyika%2C+J">James Manyika</a>, <a href="/search/cs?searchtype=author&query=Ngo%2C+H">Helen Ngo</a>, <a href="/search/cs?searchtype=author&query=Niebles%2C+J+C">Juan Carlos Niebles</a>, <a href="/search/cs?searchtype=author&query=Parli%2C+V">Vanessa Parli</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Wald%2C+R">Russell Wald</a>, <a href="/search/cs?searchtype=author&query=Clark%2C+J">Jack Clark</a>, <a href="/search/cs?searchtype=author&query=Perrault%2C+R">Raymond Perrault</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2310.03715v1-abstract-short" style="display: inline;"> Welcome to the sixth edition of the AI Index Report. This year, the report introduces more original data than any previous edition, including a new chapter on AI public opinion, a more thorough technical performance chapter, original analysis about large language and multimodal models, detailed trends in global AI legislation records, a study of the environmental impact of AI systems, and more. Th… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2310.03715v1-abstract-full').style.display = 'inline'; document.getElementById('2310.03715v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2310.03715v1-abstract-full" style="display: none;"> Welcome to the sixth edition of the AI Index Report. This year, the report introduces more original data than any previous edition, including a new chapter on AI public opinion, a more thorough technical performance chapter, original analysis about large language and multimodal models, detailed trends in global AI legislation records, a study of the environmental impact of AI systems, and more. The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Our mission is to provide unbiased, rigorously vetted, broadly sourced data in order for policymakers, researchers, executives, journalists, and the general public to develop a more thorough and nuanced understanding of the complex field of AI. The report aims to be the world's most credible and authoritative source for data and insights about AI. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2310.03715v1-abstract-full').style.display = 'none'; document.getElementById('2310.03715v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 5 October, 2023; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2023. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2307.06908">arXiv:2307.06908</a> <span> [<a href="https://arxiv.org/pdf/2307.06908">pdf</a>, <a href="https://arxiv.org/format/2307.06908">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Generating Benchmarks for Factuality Evaluation of Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Muhlgay%2C+D">Dor Muhlgay</a>, <a href="/search/cs?searchtype=author&query=Ram%2C+O">Ori Ram</a>, <a href="/search/cs?searchtype=author&query=Magar%2C+I">Inbal Magar</a>, <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Ratner%2C+N">Nir Ratner</a>, <a href="/search/cs?searchtype=author&query=Belinkov%2C+Y">Yonatan Belinkov</a>, <a href="/search/cs?searchtype=author&query=Abend%2C+O">Omri Abend</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2307.06908v2-abstract-short" style="display: inline;"> Before deploying a language model (LM) within a given domain, it is important to measure its tendency to generate factually incorrect information in that domain. Existing methods for factuality evaluation of LLM generation focus on facts sampled from the LM itself, and thus do not control the set of evaluated facts and might under-represent domain specific or rare facts. We propose FACTOR: Factual… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2307.06908v2-abstract-full').style.display = 'inline'; document.getElementById('2307.06908v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2307.06908v2-abstract-full" style="display: none;"> Before deploying a language model (LM) within a given domain, it is important to measure its tendency to generate factually incorrect information in that domain. Existing methods for factuality evaluation of LLM generation focus on facts sampled from the LM itself, and thus do not control the set of evaluated facts and might under-represent domain specific or rare facts. We propose FACTOR: Factual Assessment via Corpus TransfORmation, a scalable approach for evaluating LM factuality. FACTOR automatically transforms a factual corpus of interest into a benchmark evaluating an LM's propensity to generate true facts from the corpus vs. similar but incorrect statements. We use our framework to create three benchmarks: Wiki-FACTOR, News-FACTOR and Expert-FACTOR. We show that: (i) our benchmark scores increase with model size and improve when the LM is augmented with retrieval; (ii) benchmark score and perplexity do not always agree on model ranking; (iii) when perplexity and benchmark score disagree, the latter better reflects factuality in open-ended generation, as measured by human annotators. We make our data and code publicly available in https://github.com/AI21Labs/factor. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2307.06908v2-abstract-full').style.display = 'none'; document.getElementById('2307.06908v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 4 February, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 13 July, 2023; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2023. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2305.20010">arXiv:2305.20010</a> <span> [<a href="https://arxiv.org/pdf/2305.20010">pdf</a>, <a href="https://arxiv.org/format/2305.20010">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</span> </div> </div> <p class="title is-5 mathjax"> Human or Not? A Gamified Approach to the Turing Test </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jannai%2C+D">Daniel Jannai</a>, <a href="/search/cs?searchtype=author&query=Meron%2C+A">Amos Meron</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2305.20010v1-abstract-short" style="display: inline;"> We present "Human or Not?", an online game inspired by the Turing test, that measures the capability of AI chatbots to mimic humans in dialog, and of humans to tell bots from other humans. Over the course of a month, the game was played by over 1.5 million users who engaged in anonymous two-minute chat sessions with either another human or an AI language model which was prompted to behave like hum… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2305.20010v1-abstract-full').style.display = 'inline'; document.getElementById('2305.20010v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2305.20010v1-abstract-full" style="display: none;"> We present "Human or Not?", an online game inspired by the Turing test, that measures the capability of AI chatbots to mimic humans in dialog, and of humans to tell bots from other humans. Over the course of a month, the game was played by over 1.5 million users who engaged in anonymous two-minute chat sessions with either another human or an AI language model which was prompted to behave like humans. The task of the players was to correctly guess whether they spoke to a person or to an AI. This largest scale Turing-style test conducted to date revealed some interesting facts. For example, overall users guessed the identity of their partners correctly in only 68% of the games. In the subset of the games in which users faced an AI bot, users had even lower correct guess rates of 60% (that is, not much higher than chance). This white paper details the development, deployment, and results of this unique experiment. While this experiment calls for many extensions and refinements, these findings already begin to shed light on the inevitable near future which will commingle humans and AI. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2305.20010v1-abstract-full').style.display = 'none'; document.getElementById('2305.20010v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 31 May, 2023; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2023. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">11 pages, 6 figures</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">MSC Class:</span> 68T50 <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.2.7 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2302.00083">arXiv:2302.00083</a> <span> [<a href="https://arxiv.org/pdf/2302.00083">pdf</a>, <a href="https://arxiv.org/format/2302.00083">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Information Retrieval">cs.IR</span> </div> </div> <p class="title is-5 mathjax"> In-Context Retrieval-Augmented Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ram%2C+O">Ori Ram</a>, <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Dalmedigos%2C+I">Itay Dalmedigos</a>, <a href="/search/cs?searchtype=author&query=Muhlgay%2C+D">Dor Muhlgay</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2302.00083v3-abstract-short" style="display: inline;"> Retrieval-Augmented Language Modeling (RALM) methods, which condition a language model (LM) on relevant documents from a grounding corpus during generation, were shown to significantly improve language modeling performance. In addition, they can mitigate the problem of factually inaccurate text generation and provide natural source attribution mechanism. Existing RALM approaches focus on modifying… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2302.00083v3-abstract-full').style.display = 'inline'; document.getElementById('2302.00083v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2302.00083v3-abstract-full" style="display: none;"> Retrieval-Augmented Language Modeling (RALM) methods, which condition a language model (LM) on relevant documents from a grounding corpus during generation, were shown to significantly improve language modeling performance. In addition, they can mitigate the problem of factually inaccurate text generation and provide natural source attribution mechanism. Existing RALM approaches focus on modifying the LM architecture in order to facilitate the incorporation of external information, significantly complicating deployment. This paper considers a simple alternative, which we dub In-Context RALM: leaving the LM architecture unchanged and prepending grounding documents to the input, without any further training of the LM. We show that In-Context RALM that builds on off-the-shelf general purpose retrievers provides surprisingly large LM gains across model sizes and diverse corpora. We also demonstrate that the document retrieval and ranking mechanism can be specialized to the RALM setting to further boost performance. We conclude that In-Context RALM has considerable potential to increase the prevalence of LM grounding, particularly in settings where a pretrained LM must be used without modification or even via API access. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2302.00083v3-abstract-full').style.display = 'none'; document.getElementById('2302.00083v3-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 1 August, 2023; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 31 January, 2023; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2023. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted for publication in Transactions of the Association for Computational Linguistics (TACL). pre-MIT Press publication version</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2212.10947">arXiv:2212.10947</a> <span> [<a href="https://arxiv.org/pdf/2212.10947">pdf</a>, <a href="https://arxiv.org/format/2212.10947">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> Parallel Context Windows for Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ratner%2C+N">Nir Ratner</a>, <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Belinkov%2C+Y">Yonatan Belinkov</a>, <a href="/search/cs?searchtype=author&query=Ram%2C+O">Ori Ram</a>, <a href="/search/cs?searchtype=author&query=Magar%2C+I">Inbal Magar</a>, <a href="/search/cs?searchtype=author&query=Abend%2C+O">Omri Abend</a>, <a href="/search/cs?searchtype=author&query=Karpas%2C+E">Ehud Karpas</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2212.10947v3-abstract-short" style="display: inline;"> When applied to processing long text, Large Language Models (LLMs) are limited by their context window. Existing efforts to address this limitation involve training specialized architectures, and cannot be easily applied to off-the-shelf LLMs. We present Parallel Context Windows (PCW), a method that alleviates the context window restriction for any off-the-shelf LLM without further training. The k… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2212.10947v3-abstract-full').style.display = 'inline'; document.getElementById('2212.10947v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2212.10947v3-abstract-full" style="display: none;"> When applied to processing long text, Large Language Models (LLMs) are limited by their context window. Existing efforts to address this limitation involve training specialized architectures, and cannot be easily applied to off-the-shelf LLMs. We present Parallel Context Windows (PCW), a method that alleviates the context window restriction for any off-the-shelf LLM without further training. The key to the approach is to carve a long context into chunks (``windows''), restrict the attention mechanism to apply only within each window, and re-use the positional embeddings across the windows. Our main results test the PCW approach on in-context learning with models that range in size between 750 million and 178 billion parameters, and show substantial improvements for tasks with diverse input and output spaces. We show additional benefits in other settings where long context windows may be beneficial: multi-hop questions and retrieval-augmented question answering with multiple retrieved documents. Our results highlight Parallel Context Windows as a promising method for applying off-the-shelf LLMs in a range of settings that require long text sequences. We make our code publicly available at https://github.com/ai21labs/parallel-context-windows. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2212.10947v3-abstract-full').style.display = 'none'; document.getElementById('2212.10947v3-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 1 August, 2023; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 21 December, 2022; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> December 2022. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">The 61st Annual Meeting of the Association for Computational Linguistics (ACL 2023)</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2205.03468">arXiv:2205.03468</a> <span> [<a href="https://arxiv.org/pdf/2205.03468">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> The AI Index 2022 Annual Report </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+D">Daniel Zhang</a>, <a href="/search/cs?searchtype=author&query=Maslej%2C+N">Nestor Maslej</a>, <a href="/search/cs?searchtype=author&query=Brynjolfsson%2C+E">Erik Brynjolfsson</a>, <a href="/search/cs?searchtype=author&query=Etchemendy%2C+J">John Etchemendy</a>, <a href="/search/cs?searchtype=author&query=Lyons%2C+T">Terah Lyons</a>, <a href="/search/cs?searchtype=author&query=Manyika%2C+J">James Manyika</a>, <a href="/search/cs?searchtype=author&query=Ngo%2C+H">Helen Ngo</a>, <a href="/search/cs?searchtype=author&query=Niebles%2C+J+C">Juan Carlos Niebles</a>, <a href="/search/cs?searchtype=author&query=Sellitto%2C+M">Michael Sellitto</a>, <a href="/search/cs?searchtype=author&query=Sakhaee%2C+E">Ellie Sakhaee</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Clark%2C+J">Jack Clark</a>, <a href="/search/cs?searchtype=author&query=Perrault%2C+R">Raymond Perrault</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2205.03468v1-abstract-short" style="display: inline;"> Welcome to the fifth edition of the AI Index Report! The latest edition includes data from a broad set of academic, private, and nonprofit organizations as well as more self-collected data and original analysis than any previous editions, including an expanded technical performance chapter, a new survey of robotics researchers around the world, data on global AI legislation records in 25 countries… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2205.03468v1-abstract-full').style.display = 'inline'; document.getElementById('2205.03468v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2205.03468v1-abstract-full" style="display: none;"> Welcome to the fifth edition of the AI Index Report! The latest edition includes data from a broad set of academic, private, and nonprofit organizations as well as more self-collected data and original analysis than any previous editions, including an expanded technical performance chapter, a new survey of robotics researchers around the world, data on global AI legislation records in 25 countries, and a new chapter with an in-depth analysis of technical AI ethics metrics. The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop a more thorough and nuanced understanding of the complex field of AI. The report aims to be the world's most credible and authoritative source for data and insights about AI. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2205.03468v1-abstract-full').style.display = 'none'; document.getElementById('2205.03468v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 2 May, 2022; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2022. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2205.00445">arXiv:2205.00445</a> <span> [<a href="https://arxiv.org/pdf/2205.00445">pdf</a>, <a href="https://arxiv.org/format/2205.00445">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> MRKL Systems: A modular, neuro-symbolic architecture that combines large language models, external knowledge sources and discrete reasoning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Karpas%2C+E">Ehud Karpas</a>, <a href="/search/cs?searchtype=author&query=Abend%2C+O">Omri Abend</a>, <a href="/search/cs?searchtype=author&query=Belinkov%2C+Y">Yonatan Belinkov</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Lieber%2C+O">Opher Lieber</a>, <a href="/search/cs?searchtype=author&query=Ratner%2C+N">Nir Ratner</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Bata%2C+H">Hofit Bata</a>, <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Muhlgay%2C+D">Dor Muhlgay</a>, <a href="/search/cs?searchtype=author&query=Rozen%2C+N">Noam Rozen</a>, <a href="/search/cs?searchtype=author&query=Schwartz%2C+E">Erez Schwartz</a>, <a href="/search/cs?searchtype=author&query=Shachaf%2C+G">Gal Shachaf</a>, <a href="/search/cs?searchtype=author&query=Shalev-Shwartz%2C+S">Shai Shalev-Shwartz</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Tenenholtz%2C+M">Moshe Tenenholtz</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2205.00445v1-abstract-short" style="display: inline;"> Huge language models (LMs) have ushered in a new era for AI, serving as a gateway to natural-language-based knowledge tasks. Although an essential element of modern AI, LMs are also inherently limited in a number of ways. We discuss these limitations and how they can be avoided by adopting a systems approach. Conceptualizing the challenge as one that involves knowledge and reasoning in addition to… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2205.00445v1-abstract-full').style.display = 'inline'; document.getElementById('2205.00445v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2205.00445v1-abstract-full" style="display: none;"> Huge language models (LMs) have ushered in a new era for AI, serving as a gateway to natural-language-based knowledge tasks. Although an essential element of modern AI, LMs are also inherently limited in a number of ways. We discuss these limitations and how they can be avoided by adopting a systems approach. Conceptualizing the challenge as one that involves knowledge and reasoning in addition to linguistic processing, we define a flexible architecture with multiple neural models, complemented by discrete knowledge and reasoning modules. We describe this neuro-symbolic architecture, dubbed the Modular Reasoning, Knowledge and Language (MRKL, pronounced "miracle") system, some of the technical challenges in implementing it, and Jurassic-X, AI21 Labs' MRKL system implementation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2205.00445v1-abstract-full').style.display = 'none'; document.getElementById('2205.00445v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 1 May, 2022; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2022. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2204.10019">arXiv:2204.10019</a> <span> [<a href="https://arxiv.org/pdf/2204.10019">pdf</a>, <a href="https://arxiv.org/format/2204.10019">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Standing on the Shoulders of Giant Frozen Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Dalmedigos%2C+I">Itay Dalmedigos</a>, <a href="/search/cs?searchtype=author&query=Ram%2C+O">Ori Ram</a>, <a href="/search/cs?searchtype=author&query=Zeldes%2C+Y">Yoel Zeldes</a>, <a href="/search/cs?searchtype=author&query=Jannai%2C+D">Daniel Jannai</a>, <a href="/search/cs?searchtype=author&query=Muhlgay%2C+D">Dor Muhlgay</a>, <a href="/search/cs?searchtype=author&query=Osin%2C+Y">Yoni Osin</a>, <a href="/search/cs?searchtype=author&query=Lieber%2C+O">Opher Lieber</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Shalev-Shwartz%2C+S">Shai Shalev-Shwartz</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2204.10019v1-abstract-short" style="display: inline;"> Huge pretrained language models (LMs) have demonstrated surprisingly good zero-shot capabilities on a wide variety of tasks. This gives rise to the appealing vision of a single, versatile model with a wide range of functionalities across disparate applications. However, current leading techniques for leveraging a "frozen" LM -- i.e., leaving its weights untouched -- still often underperform fine-t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2204.10019v1-abstract-full').style.display = 'inline'; document.getElementById('2204.10019v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2204.10019v1-abstract-full" style="display: none;"> Huge pretrained language models (LMs) have demonstrated surprisingly good zero-shot capabilities on a wide variety of tasks. This gives rise to the appealing vision of a single, versatile model with a wide range of functionalities across disparate applications. However, current leading techniques for leveraging a "frozen" LM -- i.e., leaving its weights untouched -- still often underperform fine-tuning approaches which modify these weights in a task-dependent way. Those, in turn, suffer forgetfulness and compromise versatility, suggesting a tradeoff between performance and versatility. The main message of this paper is that current frozen-model techniques such as prompt tuning are only the tip of the iceberg, and more powerful methods for leveraging frozen LMs can do just as well as fine tuning in challenging domains without sacrificing the underlying model's versatility. To demonstrate this, we introduce three novel methods for leveraging frozen models: input-dependent prompt tuning, frozen readers, and recursive LMs, each of which vastly improves on current frozen-model approaches. Indeed, some of our methods even outperform fine-tuning approaches in domains currently dominated by the latter. The computational cost of each method is higher than that of existing frozen model methods, but still negligible relative to a single pass through a huge frozen LM. Each of these methods constitutes a meaningful contribution in its own right, but by presenting these contributions together we aim to convince the reader of a broader message that goes beyond the details of any given method: that frozen models have untapped potential and that fine-tuning is often unnecessary. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2204.10019v1-abstract-full').style.display = 'none'; document.getElementById('2204.10019v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 21 April, 2022; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> April 2022. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2105.01904">arXiv:2105.01904</a> <span> [<a href="https://arxiv.org/pdf/2105.01904">pdf</a>, <a href="https://arxiv.org/format/2105.01904">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Solving Sokoban with forward-backward reinforcement learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yaron Shoham</a>, <a href="/search/cs?searchtype=author&query=Elidan%2C+G">Gal Elidan</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2105.01904v2-abstract-short" style="display: inline;"> Despite seminal advances in reinforcement learning in recent years, many domains where the rewards are sparse, e.g. given only at task completion, remain quite challenging. In such cases, it can be beneficial to tackle the task both from its beginning and end, and make the two ends meet. Existing approaches that do so, however, are not effective in the common scenario where the strategy needed nea… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2105.01904v2-abstract-full').style.display = 'inline'; document.getElementById('2105.01904v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2105.01904v2-abstract-full" style="display: none;"> Despite seminal advances in reinforcement learning in recent years, many domains where the rewards are sparse, e.g. given only at task completion, remain quite challenging. In such cases, it can be beneficial to tackle the task both from its beginning and end, and make the two ends meet. Existing approaches that do so, however, are not effective in the common scenario where the strategy needed near the end goal is very different from the one that is effective earlier on. In this work we propose a novel RL approach for such settings. In short, we first train a backward-looking agent with a simple relaxed goal, and then augment the state representation of the forward-looking agent with straightforward hint features. This allows the learned forward agent to leverage information from backward plans, without mimicking their policy. We demonstrate the efficacy of our approach on the challenging game of Sokoban, where we substantially surpass learned solvers that generalize across levels, and are competitive with SOTA performance of the best highly-crafted systems. Impressively, we achieve these results while learning from a small number of practice levels and using simple RL techniques. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2105.01904v2-abstract-full').style.display = 'none'; document.getElementById('2105.01904v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 22 May, 2021; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 5 May, 2021; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2021. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">To be published in SoCS 2021</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.2.6; I.2.8 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2103.06312">arXiv:2103.06312</a> <span> [<a href="https://arxiv.org/pdf/2103.06312">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="General Literature">cs.GL</span> </div> </div> <p class="title is-5 mathjax"> The AI Index 2021 Annual Report </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+D">Daniel Zhang</a>, <a href="/search/cs?searchtype=author&query=Mishra%2C+S">Saurabh Mishra</a>, <a href="/search/cs?searchtype=author&query=Brynjolfsson%2C+E">Erik Brynjolfsson</a>, <a href="/search/cs?searchtype=author&query=Etchemendy%2C+J">John Etchemendy</a>, <a href="/search/cs?searchtype=author&query=Ganguli%2C+D">Deep Ganguli</a>, <a href="/search/cs?searchtype=author&query=Grosz%2C+B">Barbara Grosz</a>, <a href="/search/cs?searchtype=author&query=Lyons%2C+T">Terah Lyons</a>, <a href="/search/cs?searchtype=author&query=Manyika%2C+J">James Manyika</a>, <a href="/search/cs?searchtype=author&query=Niebles%2C+J+C">Juan Carlos Niebles</a>, <a href="/search/cs?searchtype=author&query=Sellitto%2C+M">Michael Sellitto</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Clark%2C+J">Jack Clark</a>, <a href="/search/cs?searchtype=author&query=Perrault%2C+R">Raymond Perrault</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2103.06312v1-abstract-short" style="display: inline;"> Welcome to the fourth edition of the AI Index Report. This year we significantly expanded the amount of data available in the report, worked with a broader set of external organizations to calibrate our data, and deepened our connections with the Stanford Institute for Human-Centered Artificial Intelligence (HAI). The AI Index Report tracks, collates, distills, and visualizes data related to artif… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2103.06312v1-abstract-full').style.display = 'inline'; document.getElementById('2103.06312v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2103.06312v1-abstract-full" style="display: none;"> Welcome to the fourth edition of the AI Index Report. This year we significantly expanded the amount of data available in the report, worked with a broader set of external organizations to calibrate our data, and deepened our connections with the Stanford Institute for Human-Centered Artificial Intelligence (HAI). The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop intuitions about the complex field of AI. The report aims to be the most credible and authoritative source for data and insights about AI in the world. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2103.06312v1-abstract-full').style.display = 'none'; document.getElementById('2103.06312v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 8 March, 2021; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2021. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2010.01825">arXiv:2010.01825</a> <span> [<a href="https://arxiv.org/pdf/2010.01825">pdf</a>, <a href="https://arxiv.org/format/2010.01825">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">stat.ML</span> </div> </div> <p class="title is-5 mathjax"> PMI-Masking: Principled masking of correlated spans </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Lieber%2C+O">Opher Lieber</a>, <a href="/search/cs?searchtype=author&query=Abend%2C+O">Omri Abend</a>, <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Tennenholtz%2C+M">Moshe Tennenholtz</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2010.01825v1-abstract-short" style="display: inline;"> Masking tokens uniformly at random constitutes a common flaw in the pretraining of Masked Language Models (MLMs) such as BERT. We show that such uniform masking allows an MLM to minimize its training objective by latching onto shallow local signals, leading to pretraining inefficiency and suboptimal downstream performance. To address this flaw, we propose PMI-Masking, a principled masking strategy… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2010.01825v1-abstract-full').style.display = 'inline'; document.getElementById('2010.01825v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2010.01825v1-abstract-full" style="display: none;"> Masking tokens uniformly at random constitutes a common flaw in the pretraining of Masked Language Models (MLMs) such as BERT. We show that such uniform masking allows an MLM to minimize its training objective by latching onto shallow local signals, leading to pretraining inefficiency and suboptimal downstream performance. To address this flaw, we propose PMI-Masking, a principled masking strategy based on the concept of Pointwise Mutual Information (PMI), which jointly masks a token n-gram if it exhibits high collocation over the corpus. PMI-Masking motivates, unifies, and improves upon prior more heuristic approaches that attempt to address the drawback of random uniform token masking, such as whole-word masking, entity/phrase masking, and random-span masking. Specifically, we show experimentally that PMI-Masking reaches the performance of prior masking approaches in half the training time, and consistently improves performance at the end of training. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2010.01825v1-abstract-full').style.display = 'none'; document.getElementById('2010.01825v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 5 October, 2020; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2020. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2004.08900">arXiv:2004.08900</a> <span> [<a href="https://arxiv.org/pdf/2004.08900">pdf</a>, <a href="https://arxiv.org/format/2004.08900">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Neural and Evolutionary Computing">cs.NE</span> </div> </div> <p class="title is-5 mathjax"> The Cost of Training NLP Models: A Concise Overview </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sharir%2C+O">Or Sharir</a>, <a href="/search/cs?searchtype=author&query=Peleg%2C+B">Barak Peleg</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2004.08900v1-abstract-short" style="display: inline;"> We review the cost of training large-scale language models, and the drivers of these costs. The intended audience includes engineers and scientists budgeting their model-training experiments, as well as non-practitioners trying to make sense of the economics of modern-day Natural Language Processing (NLP). </span> <span class="abstract-full has-text-grey-dark mathjax" id="2004.08900v1-abstract-full" style="display: none;"> We review the cost of training large-scale language models, and the drivers of these costs. The intended audience includes engineers and scientists budgeting their model-training experiments, as well as non-practitioners trying to make sense of the economics of modern-day Natural Language Processing (NLP). <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2004.08900v1-abstract-full').style.display = 'none'; document.getElementById('2004.08900v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 19 April, 2020; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> April 2020. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/1908.05646">arXiv:1908.05646</a> <span> [<a href="https://arxiv.org/pdf/1908.05646">pdf</a>, <a href="https://arxiv.org/format/1908.05646">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> SenseBERT: Driving Some Sense into BERT </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Levine%2C+Y">Yoav Levine</a>, <a href="/search/cs?searchtype=author&query=Lenz%2C+B">Barak Lenz</a>, <a href="/search/cs?searchtype=author&query=Dagan%2C+O">Or Dagan</a>, <a href="/search/cs?searchtype=author&query=Ram%2C+O">Ori Ram</a>, <a href="/search/cs?searchtype=author&query=Padnos%2C+D">Dan Padnos</a>, <a href="/search/cs?searchtype=author&query=Sharir%2C+O">Or Sharir</a>, <a href="/search/cs?searchtype=author&query=Shalev-Shwartz%2C+S">Shai Shalev-Shwartz</a>, <a href="/search/cs?searchtype=author&query=Shashua%2C+A">Amnon Shashua</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="1908.05646v2-abstract-short" style="display: inline;"> The ability to learn from large unlabeled corpora has allowed neural language models to advance the frontier in natural language understanding. However, existing self-supervision techniques operate at the word form level, which serves as a surrogate for the underlying semantic content. This paper proposes a method to employ weak-supervision directly at the word sense level. Our model, named SenseB… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1908.05646v2-abstract-full').style.display = 'inline'; document.getElementById('1908.05646v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="1908.05646v2-abstract-full" style="display: none;"> The ability to learn from large unlabeled corpora has allowed neural language models to advance the frontier in natural language understanding. However, existing self-supervision techniques operate at the word form level, which serves as a surrogate for the underlying semantic content. This paper proposes a method to employ weak-supervision directly at the word sense level. Our model, named SenseBERT, is pre-trained to predict not only the masked words but also their WordNet supersenses. Accordingly, we attain a lexical-semantic level language model, without the use of human annotation. SenseBERT achieves significantly improved lexical understanding, as we demonstrate by experimenting on SemEval Word Sense Disambiguation, and by attaining a state of the art result on the Word in Context task. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1908.05646v2-abstract-full').style.display = 'none'; document.getElementById('1908.05646v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 18 May, 2020; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 15 August, 2019; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2019. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted to ACL 2020</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/1405.7751">arXiv:1405.7751</a> <span> [<a href="https://arxiv.org/pdf/1405.7751">pdf</a>, <a href="https://arxiv.org/ps/1405.7751">ps</a>, <a href="https://arxiv.org/format/1405.7751">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> </div> </div> <p class="title is-5 mathjax"> Stable Invitations </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lee%2C+H">Hooyeon Lee</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="1405.7751v1-abstract-short" style="display: inline;"> We consider the situation in which an organizer is trying to convene an event, and needs to choose a subset of agents to be invited. Agents have preferences over how many attendees should be at the event and possibly also who the attendees should be. This induces a stability requirement: All invited agents should prefer attending to not attending, and all the other agents should not regret being n… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1405.7751v1-abstract-full').style.display = 'inline'; document.getElementById('1405.7751v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="1405.7751v1-abstract-full" style="display: none;"> We consider the situation in which an organizer is trying to convene an event, and needs to choose a subset of agents to be invited. Agents have preferences over how many attendees should be at the event and possibly also who the attendees should be. This induces a stability requirement: All invited agents should prefer attending to not attending, and all the other agents should not regret being not invited. The organizer's objective is to find the invitation of maximum size subject to the stability requirement. We investigate the computational complexity of finding the maximum stable invitation when all agents are truthful, as well as the mechanism design problem when agents may strategically misreport their preferences. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1405.7751v1-abstract-full').style.display = 'none'; document.getElementById('1405.7751v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 29 May, 2014; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 2014. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">To appear in COMSOC 2014</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/1302.1568">arXiv:1302.1568</a> <span> [<a href="https://arxiv.org/pdf/1302.1568">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Conditional Utility, Utility Independence, and Utility Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="1302.1568v1-abstract-short" style="display: inline;"> We introduce a new interpretation of two related notions - conditional utility and utility independence. Unlike the traditional interpretation, the new interpretation renders the notions the direct analogues of their probabilistic counterparts. To capture these notions formally, we appeal to the notion of utility distribution, introduced in previous paper. We show that utility distributions, wh… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1302.1568v1-abstract-full').style.display = 'inline'; document.getElementById('1302.1568v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="1302.1568v1-abstract-full" style="display: none;"> We introduce a new interpretation of two related notions - conditional utility and utility independence. Unlike the traditional interpretation, the new interpretation renders the notions the direct analogues of their probabilistic counterparts. To capture these notions formally, we appeal to the notion of utility distribution, introduced in previous paper. We show that utility distributions, which have a structure that is identical to that of probability distributions, can be viewed as a special case of an additive multiattribute utility functions, and show how this special case permits us to capture the novel senses of conditional utility and utility independence. Finally, we present the notion of utility networks, which do for utilities what Bayesian networks do for probabilities. Specifically, utility networks exploit the new interpretation of conditional utility and utility independence to compactly represent a utility distribution. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1302.1568v1-abstract-full').style.display = 'none'; document.getElementById('1302.1568v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 6 February, 2013; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2013. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Appears in Proceedings of the Thirteenth Conference on Uncertainty in Artificial Intelligence (UAI1997)</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Report number:</span> UAI-P-1997-PG-429-436 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/1301.6714">arXiv:1301.6714</a> <span> [<a href="https://arxiv.org/pdf/1301.6714">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Expected Utility Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=La+Mura%2C+P">Pierfrancesco La Mura</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="1301.6714v1-abstract-short" style="display: inline;"> We introduce a new class of graphical representations, expected utility networks (EUNs), and discuss some of its properties and potential applications to artificial intelligence and economic theory. In EUNs not only probabilities, but also utilities enjoy a modular representation. EUNs are undirected graphs with two types of arc, representing probability and utility dependencies respectively. The… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1301.6714v1-abstract-full').style.display = 'inline'; document.getElementById('1301.6714v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="1301.6714v1-abstract-full" style="display: none;"> We introduce a new class of graphical representations, expected utility networks (EUNs), and discuss some of its properties and potential applications to artificial intelligence and economic theory. In EUNs not only probabilities, but also utilities enjoy a modular representation. EUNs are undirected graphs with two types of arc, representing probability and utility dependencies respectively. The representation of utilities is based on a novel notion of conditional utility independence, which we introduce and discuss in the context of other existing proposals. Just as probabilistic inference involves the computation of conditional probabilities, strategic inference involves the computation of conditional expected utilities for alternative plans of action. We define a new notion of conditional expected utility (EU) independence, and show that in EUNs node separation with respect to the probability and utility subgraphs implies conditional EU independence. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1301.6714v1-abstract-full').style.display = 'none'; document.getElementById('1301.6714v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 23 January, 2013; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2013. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Appears in Proceedings of the Fifteenth Conference on Uncertainty in Artificial Intelligence (UAI1999)</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Report number:</span> UAI-P-1999-PG-366-373 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/1301.0595">arXiv:1301.0595</a> <span> [<a href="https://arxiv.org/pdf/1301.0595">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> </div> </div> <p class="title is-5 mathjax"> Mechanism Design with Execution Uncertainty </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Porter%2C+R">Ryan Porter</a>, <a href="/search/cs?searchtype=author&query=Ronen%2C+A">Amir Ronen</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a>, <a href="/search/cs?searchtype=author&query=Tennenholtz%2C+M">Moshe Tennenholtz</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="1301.0595v1-abstract-short" style="display: inline;"> We introduce the notion of fault tolerant mechanism design, which extends the standard game theoretic framework of mechanism design to allow for uncertainty about execution. Specifically, we define the problem of task allocation in which the private information of the agents is not only their costs to attempt the tasks, but also their probabilities of failure. For several different instances of th… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1301.0595v1-abstract-full').style.display = 'inline'; document.getElementById('1301.0595v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="1301.0595v1-abstract-full" style="display: none;"> We introduce the notion of fault tolerant mechanism design, which extends the standard game theoretic framework of mechanism design to allow for uncertainty about execution. Specifically, we define the problem of task allocation in which the private information of the agents is not only their costs to attempt the tasks, but also their probabilities of failure. For several different instances of this setting we present technical results, including positive ones in the form of mechanisms that are incentive compatible, individually rational and efficient, and negative ones in the form of impossibility theorems. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('1301.0595v1-abstract-full').style.display = 'none'; document.getElementById('1301.0595v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 12 December, 2012; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2013. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Appears in Proceedings of the Eighteenth Conference on Uncertainty in Artificial Intelligence (UAI2002)</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Report number:</span> UAI-P-2002-PG-414-421 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/cs/0202017">arXiv:cs/0202017</a> <span> [<a href="https://arxiv.org/pdf/cs/0202017">pdf</a>, <a href="https://arxiv.org/ps/cs/0202017">ps</a>, <a href="https://arxiv.org/format/cs/0202017">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> </div> </div> <p class="title is-5 mathjax"> Truth Revelation in Approximately Efficient Combinatorial Auctions </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lehmann%2C+D">Daniel Lehmann</a>, <a href="/search/cs?searchtype=author&query=O%27Callaghan%2C+L+I">Liadan Ita O'Callaghan</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="cs/0202017v1-abstract-short" style="display: inline;"> Some important classical mechanisms considered in Microeconomics and Game Theory require the solution of a difficult optimization problem. This is true of mechanisms for combinatorial auctions, which have in recent years assumed practical importance, and in particular of the gold standard for combinatorial auctions, the Generalized Vickrey Auction (GVA). Traditional analysis of these mechanisms… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/0202017v1-abstract-full').style.display = 'inline'; document.getElementById('cs/0202017v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="cs/0202017v1-abstract-full" style="display: none;"> Some important classical mechanisms considered in Microeconomics and Game Theory require the solution of a difficult optimization problem. This is true of mechanisms for combinatorial auctions, which have in recent years assumed practical importance, and in particular of the gold standard for combinatorial auctions, the Generalized Vickrey Auction (GVA). Traditional analysis of these mechanisms - in particular, their truth revelation properties - assumes that the optimization problems are solved precisely. In reality, these optimization problems can usually be solved only in an approximate fashion. We investigate the impact on such mechanisms of replacing exact solutions by approximate ones. Specifically, we look at a particular greedy optimization method. We show that the GVA payment scheme does not provide for a truth revealing mechanism. We introduce another scheme that does guarantee truthfulness for a restricted class of players. We demonstrate the latter property by identifying natural properties for combinatorial auctions and showing that, for our restricted class of players, they imply that truthful strategies are dominant. Those properties have applicability beyond the specific auction studied. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/0202017v1-abstract-full').style.display = 'none'; document.getElementById('cs/0202017v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 15 February, 2002; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2002. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Submitted to a Journal. A preliminary version appeared in EC'99</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Report number:</span> Stanford University CS-TN-99-88 <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> G.1.6; I.2.8; J.4 </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Journal ref:</span> Journal of the ACM Vol. 49, No. 5, September 2002, pp. 577-602 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/cs/0201017">arXiv:cs/0201017</a> <span> [<a href="https://arxiv.org/pdf/cs/0201017">pdf</a>, <a href="https://arxiv.org/ps/cs/0201017">ps</a>, <a href="https://arxiv.org/format/cs/0201017">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Collusion in Unrepeated, First-Price Auctions with an Uncertain Number of Participants </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Leyton-Brown%2C+K">Kevin Leyton-Brown</a>, <a href="/search/cs?searchtype=author&query=Tennenholtz%2C+M">Moshe Tennenholtz</a>, <a href="/search/cs?searchtype=author&query=Bhat%2C+N">Navin Bhat</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Yoav Shoham</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="cs/0201017v2-abstract-short" style="display: inline;"> We consider the question of whether collusion among bidders (a "bidding ring") can be supported in equilibrium of unrepeated first-price auctions. Unlike previous work on the topic such as that by McAfee and McMillan [1992] and Marshall and Marx [2007], we do not assume that non-colluding agents have perfect knowledge about the number of colluding agents whose bids are suppressed by the bidding ri… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/0201017v2-abstract-full').style.display = 'inline'; document.getElementById('cs/0201017v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="cs/0201017v2-abstract-full" style="display: none;"> We consider the question of whether collusion among bidders (a "bidding ring") can be supported in equilibrium of unrepeated first-price auctions. Unlike previous work on the topic such as that by McAfee and McMillan [1992] and Marshall and Marx [2007], we do not assume that non-colluding agents have perfect knowledge about the number of colluding agents whose bids are suppressed by the bidding ring, and indeed even allow for the existence of multiple cartels. Furthermore, while we treat the association of bidders with bidding rings as exogenous, we allow bidders to make strategic decisions about whether to join bidding rings when invited. We identify a bidding ring protocol that results in an efficient allocation in Bayes{Nash equilibrium, under which non-colluding agents bid straightforwardly, and colluding agents join bidding rings when invited and truthfully declare their valuations to the ring center. We show that bidding rings benefit ring centers and all agents, both members and non-members of bidding rings, at the auctioneer's expense. The techniques we introduce in this paper may also be useful for reasoning about other problems in which agents have asymmetric information about a setting. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/0201017v2-abstract-full').style.display = 'none'; document.getElementById('cs/0201017v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 22 August, 2012; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 18 January, 2002; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2002. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Updated the version originally submitted to Arxiv in 2002</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.2.11; J.4 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/cs/9505102">arXiv:cs/9505102</a> <span> [<a href="https://arxiv.org/pdf/cs/9505102">pdf</a>, <a href="https://arxiv.org/ps/cs/9505102">ps</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Adaptive Load Balancing: A Study in Multi-Agent Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Schaerf%2C+A">A. Schaerf</a>, <a href="/search/cs?searchtype=author&query=Shoham%2C+Y">Y. Shoham</a>, <a href="/search/cs?searchtype=author&query=Tennenholtz%2C+M">M. Tennenholtz</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="cs/9505102v1-abstract-short" style="display: inline;"> We study the process of multi-agent reinforcement learning in the context of load balancing in a distributed system, without use of either central coordination or explicit communication. We first define a precise framework in which to study adaptive load balancing, important features of which are its stochastic nature and the purely local information available to individual agents. Given this fr… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/9505102v1-abstract-full').style.display = 'inline'; document.getElementById('cs/9505102v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="cs/9505102v1-abstract-full" style="display: none;"> We study the process of multi-agent reinforcement learning in the context of load balancing in a distributed system, without use of either central coordination or explicit communication. We first define a precise framework in which to study adaptive load balancing, important features of which are its stochastic nature and the purely local information available to individual agents. Given this framework, we show illuminating results on the interplay between basic adaptive behavior parameters and their effect on system efficiency. We then investigate the properties of adaptive load balancing in heterogeneous populations, and address the issue of exploration vs. exploitation in that context. Finally, we show that naive use of communication may not improve, and might even harm system efficiency. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('cs/9505102v1-abstract-full').style.display = 'none'; document.getElementById('cs/9505102v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 30 April, 1995; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> May 1995. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">See http://www.jair.org/ for any accompanying files</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Journal ref:</span> Journal of Artificial Intelligence Research, Vol 2, (1995), 475-500 </p> </li> </ol> <div class="is-hidden-tablet">  <span class="help" style="display: inline-block;"><a href="https://github.com/arXiv/arxiv-search/releases">Search v0.5.6 released 2020-02-24</a>  </span> </div> </div> </main> <footer> <div class="columns is-desktop" role="navigation" aria-label="Secondary">  <div class="column"> <div class="columns"> <div class="column"> <ul class="nav-spaced"> <li><a href="https://info.arxiv.org/about">About</a></li> <li><a href="https://info.arxiv.org/help">Help</a></li> </ul> </div> <div class="column"> <ul class="nav-spaced"> <li> <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon filter-black" role="presentation"><title>contact arXiv</title><desc>Click here to contact arXiv</desc><path d="M502.3 190.8c3.9-3.1 9.7-.2 9.7 4.7V400c0 26.5-21.5 48-48 48H48c-26.5 0-48-21.5-48-48V195.6c0-5 5.7-7.8 9.7-4.7 22.4 17.4 52.1 39.5 154.1 113.6 21.1 15.4 56.7 47.8 92.2 47.6 35.7.3 72-32.8 92.3-47.6 102-74.1 131.6-96.3 154-113.7zM256 320c23.2.4 56.6-29.2 73.4-41.4 132.7-96.3 142.8-104.7 173.4-128.7 5.8-4.5 9.2-11.5 9.2-18.9v-19c0-26.5-21.5-48-48-48H48C21.5 64 0 85.5 0 112v19c0 7.4 3.4 14.3 9.2 18.9 30.6 23.9 40.7 32.4 173.4 128.7 16.8 12.2 50.2 41.8 73.4 41.4z"/></svg> <a href="https://info.arxiv.org/help/contact.html"> Contact</a> </li> <li> <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon filter-black" role="presentation"><title>subscribe to arXiv mailings</title><desc>Click here to subscribe</desc><path d="M476 3.2L12.5 270.6c-18.1 10.4-15.8 35.6 2.2 43.2L121 358.4l287.3-253.2c5.5-4.9 13.3 2.6 8.6 8.3L176 407v80.5c0 23.6 28.5 32.9 42.5 15.8L282 426l124.6 52.2c14.2 6 30.4-2.9 33-18.2l72-432C515 7.8 493.3-6.8 476 3.2z"/></svg> <a href="https://info.arxiv.org/help/subscribe"> Subscribe</a> </li> </ul> </div> </div> </div>   <div class="column"> <div class="columns"> <div class="column"> <ul class="nav-spaced"> <li><a href="https://info.arxiv.org/help/license/index.html">Copyright</a></li> <li><a href="https://info.arxiv.org/help/policies/privacy_policy.html">Privacy Policy</a></li> </ul> </div> <div class="column sorry-app-links"> <ul class="nav-spaced"> <li><a href="https://info.arxiv.org/help/web_accessibility.html">Web Accessibility Assistance</a></li> <li> <p class="help"> <a class="a11y-main-link" href="https://status.arxiv.org" target="_blank">arXiv Operational Status <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 256 512" class="icon filter-dark_grey" role="presentation"><path d="M224.3 273l-136 136c-9.4 9.4-24.6 9.4-33.9 0l-22.6-22.6c-9.4-9.4-9.4-24.6 0-33.9l96.4-96.4-96.4-96.4c-9.4-9.4-9.4-24.6 0-33.9L54.3 103c9.4-9.4 24.6-9.4 33.9 0l136 136c9.5 9.4 9.5 24.6.1 34z"/></svg></a><br> Get status notifications via <a class="is-link" href="https://subscribe.sorryapp.com/24846f03/email/new" target="_blank"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon filter-black" role="presentation"><path d="M502.3 190.8c3.9-3.1 9.7-.2 9.7 4.7V400c0 26.5-21.5 48-48 48H48c-26.5 0-48-21.5-48-48V195.6c0-5 5.7-7.8 9.7-4.7 22.4 17.4 52.1 39.5 154.1 113.6 21.1 15.4 56.7 47.8 92.2 47.6 35.7.3 72-32.8 92.3-47.6 102-74.1 131.6-96.3 154-113.7zM256 320c23.2.4 56.6-29.2 73.4-41.4 132.7-96.3 142.8-104.7 173.4-128.7 5.8-4.5 9.2-11.5 9.2-18.9v-19c0-26.5-21.5-48-48-48H48C21.5 64 0 85.5 0 112v19c0 7.4 3.4 14.3 9.2 18.9 30.6 23.9 40.7 32.4 173.4 128.7 16.8 12.2 50.2 41.8 73.4 41.4z"/></svg>email</a> or <a class="is-link" href="https://subscribe.sorryapp.com/24846f03/slack/new" target="_blank"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512" class="icon filter-black" role="presentation"><path d="M94.12 315.1c0 25.9-21.16 47.06-47.06 47.06S0 341 0 315.1c0-25.9 21.16-47.06 47.06-47.06h47.06v47.06zm23.72 0c0-25.9 21.16-47.06 47.06-47.06s47.06 21.16 47.06 47.06v117.84c0 25.9-21.16 47.06-47.06 47.06s-47.06-21.16-47.06-47.06V315.1zm47.06-188.98c-25.9 0-47.06-21.16-47.06-47.06S139 32 164.9 32s47.06 21.16 47.06 47.06v47.06H164.9zm0 23.72c25.9 0 47.06 21.16 47.06 47.06s-21.16 47.06-47.06 47.06H47.06C21.16 243.96 0 222.8 0 196.9s21.16-47.06 47.06-47.06H164.9zm188.98 47.06c0-25.9 21.16-47.06 47.06-47.06 25.9 0 47.06 21.16 47.06 47.06s-21.16 47.06-47.06 47.06h-47.06V196.9zm-23.72 0c0 25.9-21.16 47.06-47.06 47.06-25.9 0-47.06-21.16-47.06-47.06V79.06c0-25.9 21.16-47.06 47.06-47.06 25.9 0 47.06 21.16 47.06 47.06V196.9zM283.1 385.88c25.9 0 47.06 21.16 47.06 47.06 0 25.9-21.16 47.06-47.06 47.06-25.9 0-47.06-21.16-47.06-47.06v-47.06h47.06zm0-23.72c-25.9 0-47.06-21.16-47.06-47.06 0-25.9 21.16-47.06 47.06-47.06h117.84c25.9 0 47.06 21.16 47.06 47.06 0 25.9-21.16 47.06-47.06 47.06H283.1z"/></svg>slack</a> </p> </li> </ul> </div> </div> </div>  </div> </footer> <script src="https://static.arxiv.org/static/base/1.0.0a5/js/member_acknowledgement.js"></script> </body> </html>

CINXE.COM

Search | arXiv e-print repository