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 1–50 of 357 results for author: <span class="mathjax">Zhao, K</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=Zhao%2C+K">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="Zhao, K"> </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=Zhao%2C+K&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="Zhao, K"> <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> <nav class="pagination is-small is-centered breathe-horizontal" role="navigation" aria-label="pagination"> <a href="" class="pagination-previous is-invisible">Previous </a> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=200" class="pagination-link " aria-label="Page 5" aria-current="page">5 </a> </li> <li><span class="pagination-ellipsis">…</span></li> </ul> </nav> <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/2411.11266">arXiv:2411.11266</a> <span> [<a href="https://arxiv.org/pdf/2411.11266">pdf</a>, <a href="https://arxiv.org/format/2411.11266">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"> VersaTune: Fine-Tuning Multi-Ability LLMs Efficiently </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lu%2C+K">Keer Lu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Keshi Zhao</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+Z">Zheng Liang</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+D">Da Pan</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+S">Shusen Zhang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+X">Xin Wu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+W">Weipeng Chen</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Z">Zenan Zhou</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+G">Guosheng Dong</a>, <a href="/search/cs?searchtype=author&query=Cui%2C+B">Bin Cui</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wentao Zhang</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="2411.11266v1-abstract-short" style="display: inline;"> Large Language Models (LLMs) exhibit remarkable capabilities in handling multiple tasks across domains due to their emergent properties. These capabilities are further augmented during the Supervised Fine-Tuning (SFT) phase. Despite their potential, existing work mainly focuses on domain-specific enhancements during fine-tuning, the challenge of which lies in catastrophic forgetting of knowledge a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.11266v1-abstract-full').style.display = 'inline'; document.getElementById('2411.11266v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.11266v1-abstract-full" style="display: none;"> Large Language Models (LLMs) exhibit remarkable capabilities in handling multiple tasks across domains due to their emergent properties. These capabilities are further augmented during the Supervised Fine-Tuning (SFT) phase. Despite their potential, existing work mainly focuses on domain-specific enhancements during fine-tuning, the challenge of which lies in catastrophic forgetting of knowledge across other domains. In this study, we introduce VersaTune, a novel data composition framework designed for enhancing LLMs' overall multi-ability performances during fine-tuning. We categorize knowledge into distinct domains including law, medicine, finance, science, code. We begin with detecting the distribution of domain-specific knowledge within the base model, followed by the composition of training data that aligns with the model's existing knowledge distribution. During the fine-tuning process, weights of different domains are dynamically adjusted based on their learnable potential and forgetting degree. Experimental results demonstrate that VersaTune achieves significant improvements in multi-domain performance, with a 35.21% enhancement in comprehensive multi-domain tasks. Additionally, in scenarios where specific domain optimization is required, VersaTune reduces the degradation of performance in other domains by 38.77%, without compromising the target domain's training efficacy. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.11266v1-abstract-full').style.display = 'none'; document.getElementById('2411.11266v1-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> 17 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.08380">arXiv:2411.08380</a> <span> [<a href="https://arxiv.org/pdf/2411.08380">pdf</a>, <a href="https://arxiv.org/format/2411.08380">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Video Generation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xiaofeng Wang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+F">Feng Liu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Jiayu Wang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+G">Guosheng Zhao</a>, <a href="/search/cs?searchtype=author&query=Bao%2C+X">Xiaoyi Bao</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+Z">Zheng Zhu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yingya Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xingang Wang</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="2411.08380v1-abstract-short" style="display: inline;"> Video generation has emerged as a promising tool for world simulation, leveraging visual data to replicate real-world environments. Within this context, egocentric video generation, which centers on the human perspective, holds significant potential for enhancing applications in virtual reality, augmented reality, and gaming. However, the generation of egocentric videos presents substantial challe… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.08380v1-abstract-full').style.display = 'inline'; document.getElementById('2411.08380v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.08380v1-abstract-full" style="display: none;"> Video generation has emerged as a promising tool for world simulation, leveraging visual data to replicate real-world environments. Within this context, egocentric video generation, which centers on the human perspective, holds significant potential for enhancing applications in virtual reality, augmented reality, and gaming. However, the generation of egocentric videos presents substantial challenges due to the dynamic nature of egocentric viewpoints, the intricate diversity of actions, and the complex variety of scenes encountered. Existing datasets are inadequate for addressing these challenges effectively. To bridge this gap, we present EgoVid-5M, the first high-quality dataset specifically curated for egocentric video generation. EgoVid-5M encompasses 5 million egocentric video clips and is enriched with detailed action annotations, including fine-grained kinematic control and high-level textual descriptions. To ensure the integrity and usability of the dataset, we implement a sophisticated data cleaning pipeline designed to maintain frame consistency, action coherence, and motion smoothness under egocentric conditions. Furthermore, we introduce EgoDreamer, which is capable of generating egocentric videos driven simultaneously by action descriptions and kinematic control signals. The EgoVid-5M dataset, associated action annotations, and all data cleansing metadata will be released for the advancement of research in egocentric video generation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.08380v1-abstract-full').style.display = 'none'; document.getElementById('2411.08380v1-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> 13 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 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">Project Page: https://egovid.github.io/</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.05274">arXiv:2411.05274</a> <span> [<a href="https://arxiv.org/pdf/2411.05274">pdf</a>, <a href="https://arxiv.org/format/2411.05274">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> </div> </div> <p class="title is-5 mathjax"> Distributed-Order Fractional Graph Operating Network </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xuhao Li</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+Q">Qiyu Kang</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+F">Feng Ji</a>, <a href="/search/cs?searchtype=author&query=Ding%2C+Q">Qinxu Ding</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+Y">Yanan Zhao</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+W">Wenfei Liang</a>, <a href="/search/cs?searchtype=author&query=Tay%2C+W+P">Wee Peng Tay</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="2411.05274v1-abstract-short" style="display: inline;"> We introduce the Distributed-order fRActional Graph Operating Network (DRAGON), a novel continuous Graph Neural Network (GNN) framework that incorporates distributed-order fractional calculus. Unlike traditional continuous GNNs that utilize integer-order or single fractional-order differential equations, DRAGON uses a learnable probability distribution over a range of real numbers for the derivati… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.05274v1-abstract-full').style.display = 'inline'; document.getElementById('2411.05274v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.05274v1-abstract-full" style="display: none;"> We introduce the Distributed-order fRActional Graph Operating Network (DRAGON), a novel continuous Graph Neural Network (GNN) framework that incorporates distributed-order fractional calculus. Unlike traditional continuous GNNs that utilize integer-order or single fractional-order differential equations, DRAGON uses a learnable probability distribution over a range of real numbers for the derivative orders. By allowing a flexible and learnable superposition of multiple derivative orders, our framework captures complex graph feature updating dynamics beyond the reach of conventional models. We provide a comprehensive interpretation of our framework's capability to capture intricate dynamics through the lens of a non-Markovian graph random walk with node feature updating driven by an anomalous diffusion process over the graph. Furthermore, to highlight the versatility of the DRAGON framework, we conduct empirical evaluations across a range of graph learning tasks. The results consistently demonstrate superior performance when compared to traditional continuous GNN models. The implementation code is available at \url{https://github.com/zknus/NeurIPS-2024-DRAGON}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.05274v1-abstract-full').style.display = 'none'; document.getElementById('2411.05274v1-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> 7 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.00761">arXiv:2411.00761</a> <span> [<a href="https://arxiv.org/pdf/2411.00761">pdf</a>, <a href="https://arxiv.org/format/2411.00761">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Databases">cs.DB</span> </div> </div> <p class="title is-5 mathjax"> LCP: Enhancing Scientific Data Management with Lossy Compression for Particles </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+L">Longtao Zhang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+R">Ruoyu Li</a>, <a href="/search/cs?searchtype=author&query=Ren%2C+C">Congrong Ren</a>, <a href="/search/cs?searchtype=author&query=Di%2C+S">Sheng Di</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jinyang Liu</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+J">Jiajun Huang</a>, <a href="/search/cs?searchtype=author&query=Underwood%2C+R">Robert Underwood</a>, <a href="/search/cs?searchtype=author&query=Grosset%2C+P">Pascal Grosset</a>, <a href="/search/cs?searchtype=author&query=Tao%2C+D">Dingwen Tao</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+X">Xin Liang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+H">Hanqi Guo</a>, <a href="/search/cs?searchtype=author&query=Capello%2C+F">Franck Capello</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</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="2411.00761v1-abstract-short" style="display: inline;"> Many scientific applications opt for particles instead of meshes as their basic primitives to model complex systems composed of billions of discrete entities. Such applications span a diverse array of scientific domains, including molecular dynamics, cosmology, computational fluid dynamics, and geology. The scale of the particles in those scientific applications increases substantially thanks to t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.00761v1-abstract-full').style.display = 'inline'; document.getElementById('2411.00761v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.00761v1-abstract-full" style="display: none;"> Many scientific applications opt for particles instead of meshes as their basic primitives to model complex systems composed of billions of discrete entities. Such applications span a diverse array of scientific domains, including molecular dynamics, cosmology, computational fluid dynamics, and geology. The scale of the particles in those scientific applications increases substantially thanks to the ever-increasing computational power in high-performance computing (HPC) platforms. However, the actual gains from such increases are often undercut by obstacles in data management systems related to data storage, transfer, and processing. Lossy compression has been widely recognized as a promising solution to enhance scientific data management systems regarding such challenges, although most existing compression solutions are tailored for Cartesian grids and thus have sub-optimal results on discrete particle data. In this paper, we introduce LCP, an innovative lossy compressor designed for particle datasets, offering superior compression quality and higher speed than existing compression solutions. Specifically, our contribution is threefold. (1) We propose LCP-S, an error-bound aware block-wise spatial compressor to efficiently reduce particle data size. This approach is universally applicable to particle data across various domains. (2) We develop LCP, a hybrid compression solution for multi-frame particle data, featuring dynamic method selection and parameter optimization. (3) We evaluate our solution alongside eight state-of-the-art alternatives on eight real-world particle datasets from seven distinct domains. The results demonstrate that our solution achieves up to 104% improvement in compression ratios and up to 593% increase in speed compared to the second-best option, under the same error criteria. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.00761v1-abstract-full').style.display = 'none'; document.getElementById('2411.00761v1-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 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">Accepted by SIGMOD'25</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.20487">arXiv:2410.20487</a> <span> [<a href="https://arxiv.org/pdf/2410.20487">pdf</a>, <a href="https://arxiv.org/format/2410.20487">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"> Efficient Diversity-based Experience Replay for Deep Reinforcement Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiyan Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yiming Wang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yuyang Chen</a>, <a href="/search/cs?searchtype=author&query=Niu%2C+X">Xiaoguang Niu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yan Li</a>, <a href="/search/cs?searchtype=author&query=U%2C+L+H">Leong Hou U</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="2410.20487v1-abstract-short" style="display: inline;"> Deep Reinforcement Learning (DRL) has achieved remarkable success in solving complex decision-making problems by combining the representation capabilities of deep learning with the decision-making power of reinforcement learning. However, learning in sparse reward environments remains challenging due to insufficient feedback to guide the optimization of agents, especially in real-life environments… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20487v1-abstract-full').style.display = 'inline'; document.getElementById('2410.20487v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.20487v1-abstract-full" style="display: none;"> Deep Reinforcement Learning (DRL) has achieved remarkable success in solving complex decision-making problems by combining the representation capabilities of deep learning with the decision-making power of reinforcement learning. However, learning in sparse reward environments remains challenging due to insufficient feedback to guide the optimization of agents, especially in real-life environments with high-dimensional states. To tackle this issue, experience replay is commonly introduced to enhance learning efficiency through past experiences. Nonetheless, current methods of experience replay, whether based on uniform or prioritized sampling, frequently struggle with suboptimal learning efficiency and insufficient utilization of samples. This paper proposes a novel approach, diversity-based experience replay (DBER), which leverages the deterministic point process to prioritize diverse samples in state realizations. We conducted extensive experiments on Robotic Manipulation tasks in MuJoCo, Atari games, and realistic in-door environments in Habitat. The results show that our method not only significantly improves learning efficiency but also demonstrates superior performance in sparse reward environments with high-dimensional states, providing a simple yet effective solution for this field. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20487v1-abstract-full').style.display = 'none'; document.getElementById('2410.20487v1-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> 27 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.18112">arXiv:2410.18112</a> <span> [<a href="https://arxiv.org/pdf/2410.18112">pdf</a>, <a href="https://arxiv.org/format/2410.18112">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Multiagent Systems">cs.MA</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="Robotics">cs.RO</span> </div> </div> <p class="title is-5 mathjax"> OPTIMA: Optimized Policy for Intelligent Multi-Agent Systems Enables Coordination-Aware Autonomous Vehicles </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Du%2C+R">Rui Du</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Hou%2C+J">Jinlong Hou</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Q">Qiang Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+P">Peter Zhang</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="2410.18112v1-abstract-short" style="display: inline;"> Coordination among connected and autonomous vehicles (CAVs) is advancing due to developments in control and communication technologies. However, much of the current work is based on oversimplified and unrealistic task-specific assumptions, which may introduce vulnerabilities. This is critical because CAVs not only interact with their environment but are also integral parts of it. Insufficient expl… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.18112v1-abstract-full').style.display = 'inline'; document.getElementById('2410.18112v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.18112v1-abstract-full" style="display: none;"> Coordination among connected and autonomous vehicles (CAVs) is advancing due to developments in control and communication technologies. However, much of the current work is based on oversimplified and unrealistic task-specific assumptions, which may introduce vulnerabilities. This is critical because CAVs not only interact with their environment but are also integral parts of it. Insufficient exploration can result in policies that carry latent risks, highlighting the need for methods that explore the environment both extensively and efficiently. This work introduces OPTIMA, a novel distributed reinforcement learning framework for cooperative autonomous vehicle tasks. OPTIMA alternates between thorough data sampling from environmental interactions and multi-agent reinforcement learning algorithms to optimize CAV cooperation, emphasizing both safety and efficiency. Our goal is to improve the generality and performance of CAVs in highly complex and crowded scenarios. Furthermore, the industrial-scale distributed training system easily adapts to different algorithms, reward functions, and strategies. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.18112v1-abstract-full').style.display = 'none'; document.getElementById('2410.18112v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.16888">arXiv:2410.16888</a> <span> [<a href="https://arxiv.org/pdf/2410.16888">pdf</a>, <a href="https://arxiv.org/format/2410.16888">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> </div> </div> <p class="title is-5 mathjax"> Unsupervised Time Series Anomaly Prediction with Importance-based Generative Contrastive Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Zhuang%2C+Z">Zhihao Zhuang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+C">Chenjuan Guo</a>, <a href="/search/cs?searchtype=author&query=Miao%2C+H">Hao Miao</a>, <a href="/search/cs?searchtype=author&query=Cheng%2C+Y">Yunyao Cheng</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bin Yang</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="2410.16888v1-abstract-short" style="display: inline;"> Time series anomaly prediction plays an essential role in many real-world scenarios, such as environmental prevention and prompt maintenance of cyber-physical systems. However, existing time series anomaly prediction methods mainly require supervised training with plenty of manually labeled data, which are difficult to obtain in practice. Besides, unseen anomalies can occur during inference, which… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16888v1-abstract-full').style.display = 'inline'; document.getElementById('2410.16888v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.16888v1-abstract-full" style="display: none;"> Time series anomaly prediction plays an essential role in many real-world scenarios, such as environmental prevention and prompt maintenance of cyber-physical systems. However, existing time series anomaly prediction methods mainly require supervised training with plenty of manually labeled data, which are difficult to obtain in practice. Besides, unseen anomalies can occur during inference, which could differ from the labeled training data and make these models fail to predict such new anomalies. In this paper, we study a novel problem of unsupervised time series anomaly prediction. We provide a theoretical analysis and propose Importance-based Generative Contrastive Learning (IGCL) to address the aforementioned problems. IGCL distinguishes between normal and anomaly precursors, which are generated by our anomaly precursor pattern generation module. To address the efficiency issues caused by the potential complex anomaly precursor combinations, we propose a memory bank with importance-based scores to adaptively store representative anomaly precursors and generate more complicated anomaly precursors. Extensive experiments on seven benchmark datasets show our method outperforms state-of-the-art baselines on unsupervised time series anomaly prediction problems. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16888v1-abstract-full').style.display = 'none'; document.getElementById('2410.16888v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">16 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.16663">arXiv:2410.16663</a> <span> [<a href="https://arxiv.org/pdf/2410.16663">pdf</a>, <a href="https://arxiv.org/format/2410.16663">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> </div> </div> <p class="title is-5 mathjax"> FastAttention: Extend FlashAttention2 to NPUs and Low-resource GPUs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lin%2C+H">Haoran Lin</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+X">Xianzhi Yu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Hou%2C+L">Lu Hou</a>, <a href="/search/cs?searchtype=author&query=Zhan%2C+Z">Zongyuan Zhan</a>, <a href="/search/cs?searchtype=author&query=Kamenev%2C+S">Stanislav Kamenev</a>, <a href="/search/cs?searchtype=author&query=Bao%2C+H">Han Bao</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+T">Ting Hu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+M">Mingkai Wang</a>, <a href="/search/cs?searchtype=author&query=Chang%2C+Q">Qixin Chang</a>, <a href="/search/cs?searchtype=author&query=Sui%2C+S">Siyue Sui</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+W">Weihao Sun</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+J">Jiaxin Hu</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+J">Jun Yao</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+Z">Zekun Yin</a>, <a href="/search/cs?searchtype=author&query=Qian%2C+C">Cheng Qian</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Ying Zhang</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+Y">Yinfei Pan</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Y">Yu Yang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+W">Weiguo Liu</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="2410.16663v1-abstract-short" style="display: inline;"> FlashAttention series has been widely applied in the inference of large language models (LLMs). However, FlashAttention series only supports the high-level GPU architectures, e.g., Ampere and Hopper. At present, FlashAttention series is not easily transferrable to NPUs and low-resource GPUs. Moreover, FlashAttention series is inefficient for multi- NPUs or GPUs inference scenarios. In this work, w… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16663v1-abstract-full').style.display = 'inline'; document.getElementById('2410.16663v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.16663v1-abstract-full" style="display: none;"> FlashAttention series has been widely applied in the inference of large language models (LLMs). However, FlashAttention series only supports the high-level GPU architectures, e.g., Ampere and Hopper. At present, FlashAttention series is not easily transferrable to NPUs and low-resource GPUs. Moreover, FlashAttention series is inefficient for multi- NPUs or GPUs inference scenarios. In this work, we propose FastAttention which pioneers the adaptation of FlashAttention series for NPUs and low-resource GPUs to boost LLM inference efficiency. Specifically, we take Ascend NPUs and Volta-based GPUs as representatives for designing our FastAttention. We migrate FlashAttention series to Ascend NPUs by proposing a novel two-level tiling strategy for runtime speedup, tiling-mask strategy for memory saving and the tiling-AllReduce strategy for reducing communication overhead, respectively. Besides, we adapt FlashAttention for Volta-based GPUs by redesigning the operands layout in shared memory and introducing a simple yet effective CPU-GPU cooperative strategy for efficient memory utilization. On Ascend NPUs, our FastAttention can achieve a 10.7$\times$ speedup compared to the standard attention implementation. Llama-7B within FastAttention reaches up to 5.16$\times$ higher throughput than within the standard attention. On Volta architecture GPUs, FastAttention yields 1.43$\times$ speedup compared to its equivalents in \texttt{xformers}. Pangu-38B within FastAttention brings 1.46$\times$ end-to-end speedup using FasterTransformer. Coupled with the propose CPU-GPU cooperative strategy, FastAttention supports a maximal input length of 256K on 8 V100 GPUs. All the codes will be made available soon. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16663v1-abstract-full').style.display = 'none'; document.getElementById('2410.16663v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.16516">arXiv:2410.16516</a> <span> [<a href="https://arxiv.org/pdf/2410.16516">pdf</a>, <a href="https://arxiv.org/format/2410.16516">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> </div> </div> <p class="title is-5 mathjax"> Scalability of memorization-based machine unlearning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kairan Zhao</a>, <a href="/search/cs?searchtype=author&query=Triantafillou%2C+P">Peter Triantafillou</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="2410.16516v2-abstract-short" style="display: inline;"> Machine unlearning (MUL) focuses on removing the influence of specific subsets of data (such as noisy, poisoned, or privacy-sensitive data) from pretrained models. MUL methods typically rely on specialized forms of fine-tuning. Recent research has shown that data memorization is a key characteristic defining the difficulty of MUL. As a result, novel memorization-based unlearning methods have been… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16516v2-abstract-full').style.display = 'inline'; document.getElementById('2410.16516v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.16516v2-abstract-full" style="display: none;"> Machine unlearning (MUL) focuses on removing the influence of specific subsets of data (such as noisy, poisoned, or privacy-sensitive data) from pretrained models. MUL methods typically rely on specialized forms of fine-tuning. Recent research has shown that data memorization is a key characteristic defining the difficulty of MUL. As a result, novel memorization-based unlearning methods have been developed, demonstrating exceptional performance with respect to unlearning quality, while maintaining high performance for model utility. Alas, these methods depend on knowing the memorization scores of data points and computing said scores is a notoriously time-consuming process. This in turn severely limits the scalability of these solutions and their practical impact for real-world applications. In this work, we tackle these scalability challenges of state-of-the-art memorization-based MUL algorithms using a series of memorization-score proxies. We first analyze the profiles of various proxies and then evaluate the performance of state-of-the-art (memorization-based) MUL algorithms in terms of both accuracy and privacy preservation. Our empirical results show that these proxies can introduce accuracy on par with full memorization-based unlearning while dramatically improving scalability. We view this work as an important step toward scalable and efficient machine unlearning. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16516v2-abstract-full').style.display = 'none'; document.getElementById('2410.16516v2-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 21 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">NeurIPS 2024 FITML Workshop</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.16135">arXiv:2410.16135</a> <span> [<a href="https://arxiv.org/pdf/2410.16135">pdf</a>, <a href="https://arxiv.org/format/2410.16135">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"> Beyond 2:4: exploring V:N:M sparsity for efficient transformer inference on GPUs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+T">Tao Yuan</a>, <a href="/search/cs?searchtype=author&query=Bao%2C+H">Han Bao</a>, <a href="/search/cs?searchtype=author&query=Su%2C+Z">Zhenfeng Su</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+C">Chang Gao</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+Z">Zhaofeng Sun</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+Z">Zichen Liang</a>, <a href="/search/cs?searchtype=author&query=Jing%2C+L">Liping Jing</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+J">Jianfei Chen</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="2410.16135v1-abstract-short" style="display: inline;"> To date, 2:4 sparsity has stood as the only sparse pattern that can be accelerated using sparse tensor cores on GPUs. In practice, 2:4 sparsity often possesses low actual speedups ($\leq 1.3$) and requires fixed sparse ratios, meaning that other ratios, such as 4:8, 8:16, or those exceeding 50% sparsity, do not incur any speedups on GPUs. Recent studies suggest that V:N:M sparsity is promising in… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16135v1-abstract-full').style.display = 'inline'; document.getElementById('2410.16135v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.16135v1-abstract-full" style="display: none;"> To date, 2:4 sparsity has stood as the only sparse pattern that can be accelerated using sparse tensor cores on GPUs. In practice, 2:4 sparsity often possesses low actual speedups ($\leq 1.3$) and requires fixed sparse ratios, meaning that other ratios, such as 4:8, 8:16, or those exceeding 50% sparsity, do not incur any speedups on GPUs. Recent studies suggest that V:N:M sparsity is promising in addressing these limitations of 2:4 sparsity. However, regarding accuracy, the effects of V:N:M sparsity on broader Transformer models, such as vision Transformers and large language models (LLMs), are largely unexamined. Moreover, Some specific issues related to V:N:M sparsity, such as how to select appropriate V and M values, remain unresolved. In this study, we thoroughly investigate the application of V:N:M sparsity in vision models and LLMs across multiple tasks, from pertaining to downstream tasks. We propose three key approaches to enhance the applicability and accuracy of V:N:M-sparse Transformers, including heuristic V and M selection, V:N:M-specific channel permutation, and three-staged LoRA training techniques. Experimental results show that, with our methods, the DeiT-small achieves lossless accuracy at 64:2:5 sparsity, while the DeiT-base maintains accuracy even at 64:2:8 sparsity. In addition, the fine-tuned LLama2-7B at 64:2:5 sparsity performs comparably or better than training-free 2:4 sparse alternatives on downstream tasks. More importantly, V:N:M-sparse Transformers offer a wider range of speedup-accuracy trade-offs compared to 2:4 sparsity. Overall, our exploration largely facilitates the V:N:M sparsity to act as a truly effective acceleration solution for Transformers in cost-sensitive inference scenarios. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.16135v1-abstract-full').style.display = 'none'; document.getElementById('2410.16135v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.15997">arXiv:2410.15997</a> <span> [<a href="https://arxiv.org/pdf/2410.15997">pdf</a>, <a href="https://arxiv.org/format/2410.15997">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> </div> </div> <p class="title is-5 mathjax"> MultiRC: Joint Learning for Time Series Anomaly Prediction and Detection with Multi-scale Reconstructive Contrast </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hu%2C+S">Shiyan Hu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Qiu%2C+X">Xiangfei Qiu</a>, <a href="/search/cs?searchtype=author&query=Shu%2C+Y">Yang Shu</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+J">Jilin Hu</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bin Yang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+C">Chenjuan Guo</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="2410.15997v1-abstract-short" style="display: inline;"> Many methods have been proposed for unsupervised time series anomaly detection. Despite some progress, research on predicting future anomalies is still relatively scarce. Predicting anomalies is particularly challenging due to the diverse reaction time and the lack of labeled data. To address these challenges, we propose MultiRC to integrate reconstructive and contrastive learning for joint learni… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.15997v1-abstract-full').style.display = 'inline'; document.getElementById('2410.15997v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.15997v1-abstract-full" style="display: none;"> Many methods have been proposed for unsupervised time series anomaly detection. Despite some progress, research on predicting future anomalies is still relatively scarce. Predicting anomalies is particularly challenging due to the diverse reaction time and the lack of labeled data. To address these challenges, we propose MultiRC to integrate reconstructive and contrastive learning for joint learning of anomaly prediction and detection, with multi-scale structure and adaptive dominant period mask to deal with the diverse reaction time. MultiRC also generates negative samples to provide essential training momentum for the anomaly prediction tasks and prevent model degradation. We evaluate seven benchmark datasets from different fields. For both anomaly prediction and detection tasks, MultiRC outperforms existing state-of-the-art methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.15997v1-abstract-full').style.display = 'none'; document.getElementById('2410.15997v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.13419">arXiv:2410.13419</a> <span> [<a href="https://arxiv.org/pdf/2410.13419">pdf</a>, <a href="https://arxiv.org/format/2410.13419">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Sound">cs.SD</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Multimedia">cs.MM</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</span> </div> </div> <p class="title is-5 mathjax"> MeloTrans: A Text to Symbolic Music Generation Model Following Human Composition Habit </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yutian Wang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+W">Wanyin Yang</a>, <a href="/search/cs?searchtype=author&query=Dai%2C+Z">Zhenrong Dai</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yilong Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Hui Wang</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="2410.13419v1-abstract-short" style="display: inline;"> At present, neural network models show powerful sequence prediction ability and are used in many automatic composition models. In comparison, the way humans compose music is very different from it. Composers usually start by creating musical motifs and then develop them into music through a series of rules. This process ensures that the music has a specific structure and changing pattern. However,… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.13419v1-abstract-full').style.display = 'inline'; document.getElementById('2410.13419v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.13419v1-abstract-full" style="display: none;"> At present, neural network models show powerful sequence prediction ability and are used in many automatic composition models. In comparison, the way humans compose music is very different from it. Composers usually start by creating musical motifs and then develop them into music through a series of rules. This process ensures that the music has a specific structure and changing pattern. However, it is difficult for neural network models to learn these composition rules from training data, which results in a lack of musicality and diversity in the generated music. This paper posits that integrating the learning capabilities of neural networks with human-derived knowledge may lead to better results. To archive this, we develop the POP909$\_$M dataset, the first to include labels for musical motifs and their variants, providing a basis for mimicking human compositional habits. Building on this, we propose MeloTrans, a text-to-music composition model that employs principles of motif development rules. Our experiments demonstrate that MeloTrans excels beyond existing music generation models and even surpasses Large Language Models (LLMs) like ChatGPT-4. This highlights the importance of merging human insights with neural network capabilities to achieve superior symbolic music generation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.13419v1-abstract-full').style.display = 'none'; document.getElementById('2410.13419v1-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> 17 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.12707">arXiv:2410.12707</a> <span> [<a href="https://arxiv.org/pdf/2410.12707">pdf</a>, <a href="https://arxiv.org/format/2410.12707">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</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"> FusionLLM: A Decentralized LLM Training System on Geo-distributed GPUs with Adaptive Compression </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tang%2C+Z">Zhenheng Tang</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+X">Xueze Kang</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+Y">Yiming Yin</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+X">Xinglin Pan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yuxin Wang</a>, <a href="/search/cs?searchtype=author&query=He%2C+X">Xin He</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Q">Qiang Wang</a>, <a href="/search/cs?searchtype=author&query=Zeng%2C+R">Rongfei Zeng</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiyong Zhao</a>, <a href="/search/cs?searchtype=author&query=Shi%2C+S">Shaohuai Shi</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+A+C">Amelie Chi Zhou</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=He%2C+B">Bingsheng He</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+X">Xiaowen Chu</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="2410.12707v1-abstract-short" style="display: inline;"> To alleviate hardware scarcity in training large deep neural networks (DNNs), particularly large language models (LLMs), we present FusionLLM, a decentralized training system designed and implemented for training DNNs using geo-distributed GPUs across different computing clusters or individual devices. Decentralized training faces significant challenges regarding system design and efficiency, incl… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12707v1-abstract-full').style.display = 'inline'; document.getElementById('2410.12707v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.12707v1-abstract-full" style="display: none;"> To alleviate hardware scarcity in training large deep neural networks (DNNs), particularly large language models (LLMs), we present FusionLLM, a decentralized training system designed and implemented for training DNNs using geo-distributed GPUs across different computing clusters or individual devices. Decentralized training faces significant challenges regarding system design and efficiency, including: 1) the need for remote automatic differentiation (RAD), 2) support for flexible model definitions and heterogeneous software, 3) heterogeneous hardware leading to low resource utilization or the straggler problem, and 4) slow network communication. To address these challenges, in the system design, we represent the model as a directed acyclic graph of operators (OP-DAG). Each node in the DAG represents the operator in the DNNs, while the edge represents the data dependency between operators. Based on this design, 1) users are allowed to customize any DNN without caring low-level operator implementation; 2) we enable the task scheduling with the more fine-grained sub-tasks, offering more optimization space; 3) a DAG runtime executor can implement RAD withour requiring the consistent low-level ML framework versions. To enhance system efficiency, we implement a workload estimator and design an OP-Fence scheduler to cluster devices with similar bandwidths together and partition the DAG to increase throughput. Additionally, we propose an AdaTopK compressor to adaptively compress intermediate activations and gradients at the slowest communication links. To evaluate the convergence and efficiency of our system and algorithms, we train ResNet-101 and GPT-2 on three real-world testbeds using 48 GPUs connected with 8 Mbps~10 Gbps networks. Experimental results demonstrate that our system and method can achieve 1.45 - 9.39x speedup compared to baseline methods while ensuring convergence. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12707v1-abstract-full').style.display = 'none'; document.getElementById('2410.12707v1-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> 16 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.12236">arXiv:2410.12236</a> <span> [<a href="https://arxiv.org/pdf/2410.12236">pdf</a>, <a href="https://arxiv.org/format/2410.12236">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"> Enhancing LLM Agents for Code Generation with Possibility and Pass-rate Prioritized Experience Replay </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yuyang Chen</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiyan Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yiming Wang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+M">Ming Yang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jian Zhang</a>, <a href="/search/cs?searchtype=author&query=Niu%2C+X">Xiaoguang Niu</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="2410.12236v1-abstract-short" style="display: inline;"> Nowadays transformer-based Large Language Models (LLM) for code generation tasks usually apply sampling and filtering pipelines. Due to the sparse reward problem in code generation tasks caused by one-token incorrectness, transformer-based models will sample redundant programs till they find a correct one, leading to low efficiency. To overcome the challenge, we incorporate Experience Replay (ER)… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12236v1-abstract-full').style.display = 'inline'; document.getElementById('2410.12236v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.12236v1-abstract-full" style="display: none;"> Nowadays transformer-based Large Language Models (LLM) for code generation tasks usually apply sampling and filtering pipelines. Due to the sparse reward problem in code generation tasks caused by one-token incorrectness, transformer-based models will sample redundant programs till they find a correct one, leading to low efficiency. To overcome the challenge, we incorporate Experience Replay (ER) in the fine-tuning phase, where codes and programs produced are stored and will be replayed to give the LLM agent a chance to learn from past experiences. Based on the spirit of ER, we introduce a novel approach called BTP pipeline which consists of three phases: beam search sampling, testing phase, and prioritized experience replay phase. The approach makes use of failed programs collected by code models and replays programs with high Possibility and Pass-rate Prioritized value (P2Value) from the replay buffer to improve efficiency. P2Value comprehensively considers the possibility of transformers' output and pass rate and can make use of the redundant resources caused by the problem that most programs collected by LLMs fail to pass any tests. We empirically apply our approach in several LLMs, demonstrating that it enhances their performance in code generation tasks and surpasses existing baselines. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12236v1-abstract-full').style.display = 'none'; document.getElementById('2410.12236v1-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> 16 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.09426">arXiv:2410.09426</a> <span> [<a href="https://arxiv.org/pdf/2410.09426">pdf</a>, <a href="https://arxiv.org/format/2410.09426">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"> FlatQuant: Flatness Matters for LLM Quantization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sun%2C+Y">Yuxuan Sun</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+R">Ruikang Liu</a>, <a href="/search/cs?searchtype=author&query=Bai%2C+H">Haoli Bai</a>, <a href="/search/cs?searchtype=author&query=Bao%2C+H">Han Bao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yuening Li</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+J">Jiaxin Hu</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+X">Xianzhi Yu</a>, <a href="/search/cs?searchtype=author&query=Hou%2C+L">Lu Hou</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+C">Chun Yuan</a>, <a href="/search/cs?searchtype=author&query=Jiang%2C+X">Xin Jiang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+W">Wulong Liu</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+J">Jun Yao</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="2410.09426v1-abstract-short" style="display: inline;"> Recently, quantization has been widely used for the compression and acceleration of large language models~(LLMs). Due to the outliers in LLMs, it is crucial to flatten weights and activations to minimize quantization error with the equally spaced quantization points. Prior research explores various pre-quantization transformations to suppress outliers, such as per-channel scaling and Hadamard tran… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09426v1-abstract-full').style.display = 'inline'; document.getElementById('2410.09426v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.09426v1-abstract-full" style="display: none;"> Recently, quantization has been widely used for the compression and acceleration of large language models~(LLMs). Due to the outliers in LLMs, it is crucial to flatten weights and activations to minimize quantization error with the equally spaced quantization points. Prior research explores various pre-quantization transformations to suppress outliers, such as per-channel scaling and Hadamard transformation. However, we observe that these transformed weights and activations can still remain steep and outspread. In this paper, we propose FlatQuant (Fast and Learnable Affine Transformation), a new post-training quantization approach to enhance flatness of weights and activations. Our approach identifies optimal affine transformations tailored to each linear layer, calibrated in hours via a lightweight objective. To reduce runtime overhead, we apply Kronecker decomposition to the transformation matrices, and fuse all operations in FlatQuant into a single kernel. Extensive experiments show that FlatQuant sets up a new state-of-the-art quantization benchmark. For instance, it achieves less than $\textbf{1}\%$ accuracy drop for W4A4 quantization on the LLaMA-3-70B model, surpassing SpinQuant by $\textbf{7.5}\%$. For inference latency, FlatQuant reduces the slowdown induced by pre-quantization transformation from 0.26x of QuaRot to merely $\textbf{0.07x}$, bringing up to $\textbf{2.3x}$ speedup for prefill and $\textbf{1.7x}$ speedup for decoding, respectively. Code is available at: \url{https://github.com/ruikangliu/FlatQuant}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09426v1-abstract-full').style.display = 'none'; document.getElementById('2410.09426v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">23 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.07701">arXiv:2410.07701</a> <span> [<a href="https://arxiv.org/pdf/2410.07701">pdf</a>, <a href="https://arxiv.org/format/2410.07701">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> </div> </div> <p class="title is-5 mathjax"> Autonomous Driving in Unstructured Environments: How Far Have We Come? </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Min%2C+C">Chen Min</a>, <a href="/search/cs?searchtype=author&query=Si%2C+S">Shubin Si</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xu Wang</a>, <a href="/search/cs?searchtype=author&query=Xue%2C+H">Hanzhang Xue</a>, <a href="/search/cs?searchtype=author&query=Jiang%2C+W">Weizhong Jiang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Y">Yang Liu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Juan Wang</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+Q">Qingtian Zhu</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+Q">Qi Zhu</a>, <a href="/search/cs?searchtype=author&query=Luo%2C+L">Lun Luo</a>, <a href="/search/cs?searchtype=author&query=Kong%2C+F">Fanjie Kong</a>, <a href="/search/cs?searchtype=author&query=Miao%2C+J">Jinyu Miao</a>, <a href="/search/cs?searchtype=author&query=Cai%2C+X">Xudong Cai</a>, <a href="/search/cs?searchtype=author&query=An%2C+S">Shuai An</a>, <a href="/search/cs?searchtype=author&query=Li%2C+W">Wei Li</a>, <a href="/search/cs?searchtype=author&query=Mei%2C+J">Jilin Mei</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+T">Tong Sun</a>, <a href="/search/cs?searchtype=author&query=Zhai%2C+H">Heng Zhai</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Q">Qifeng Liu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+F">Fangzhou Zhao</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+L">Liang Chen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+S">Shuai Wang</a>, <a href="/search/cs?searchtype=author&query=Shang%2C+E">Erke Shang</a>, <a href="/search/cs?searchtype=author&query=Shang%2C+L">Linzhi Shang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kunlong Zhao</a> , et al. (13 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="2410.07701v3-abstract-short" style="display: inline;"> Research on autonomous driving in unstructured outdoor environments is less advanced than in structured urban settings due to challenges like environmental diversities and scene complexity. These environments-such as rural areas and rugged terrains-pose unique obstacles that are not common in structured urban areas. Despite these difficulties, autonomous driving in unstructured outdoor environment… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.07701v3-abstract-full').style.display = 'inline'; document.getElementById('2410.07701v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.07701v3-abstract-full" style="display: none;"> Research on autonomous driving in unstructured outdoor environments is less advanced than in structured urban settings due to challenges like environmental diversities and scene complexity. These environments-such as rural areas and rugged terrains-pose unique obstacles that are not common in structured urban areas. Despite these difficulties, autonomous driving in unstructured outdoor environments is crucial for applications in agriculture, mining, and military operations. Our survey reviews over 250 papers for autonomous driving in unstructured outdoor environments, covering offline mapping, pose estimation, environmental perception, path planning, end-to-end autonomous driving, datasets, and relevant challenges. We also discuss emerging trends and future research directions. This review aims to consolidate knowledge and encourage further research for autonomous driving in unstructured environments. To support ongoing work, we maintain an active repository with up-to-date literature and open-source projects at: https://github.com/chaytonmin/Survey-Autonomous-Driving-in-Unstructured-Environments. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.07701v3-abstract-full').style.display = 'none'; document.getElementById('2410.07701v3-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 10 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">Survey paper; 38 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.06153">arXiv:2410.06153</a> <span> [<a href="https://arxiv.org/pdf/2410.06153">pdf</a>, <a href="https://arxiv.org/format/2410.06153">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"> AgentSquare: Automatic LLM Agent Search in Modular Design Space </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shang%2C+Y">Yu Shang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yu Li</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Keyu Zhao</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+L">Likai Ma</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jiahe Liu</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+F">Fengli Xu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yong Li</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="2410.06153v2-abstract-short" style="display: inline;"> Recent advancements in Large Language Models (LLMs) have led to a rapid growth of agentic systems capable of handling a wide range of complex tasks. However, current research largely relies on manual, task-specific design, limiting their adaptability to novel tasks. In this paper, we introduce a new research problem: Modularized LLM Agent Search (MoLAS). We propose a modular design space that abst… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06153v2-abstract-full').style.display = 'inline'; document.getElementById('2410.06153v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.06153v2-abstract-full" style="display: none;"> Recent advancements in Large Language Models (LLMs) have led to a rapid growth of agentic systems capable of handling a wide range of complex tasks. However, current research largely relies on manual, task-specific design, limiting their adaptability to novel tasks. In this paper, we introduce a new research problem: Modularized LLM Agent Search (MoLAS). We propose a modular design space that abstracts existing LLM agent designs into four fundamental modules with uniform IO interface: Planning, Reasoning, Tool Use, and Memory. Building on this design space, we present a novel LLM agent search framework called AgentSquare, which introduces two core mechanisms, i.e., module evolution and recombination, to efficiently search for optimized LLM agents. To further accelerate the process, we design a performance predictor that uses in-context surrogate models to skip unpromising agent designs. Extensive experiments across six benchmarks, covering the diverse scenarios of web, embodied, tool use and game applications, show that AgentSquare substantially outperforms hand-crafted agents, achieving an average performance gain of 17.2% against best-known human designs. Moreover, AgentSquare can generate interpretable design insights, enabling a deeper understanding of agentic architecture and its impact on task performance. We believe that the modular design space and AgentSquare search framework offer a platform for fully exploiting the potential of prior successful designs and consolidating the collective efforts of research community. Code repo is available at https://github.com/tsinghua-fib-lab/AgentSquare. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06153v2-abstract-full').style.display = 'none'; document.getElementById('2410.06153v2-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 8 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">26 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.05260">arXiv:2410.05260</a> <span> [<a href="https://arxiv.org/pdf/2410.05260">pdf</a>, <a href="https://arxiv.org/format/2410.05260">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Graphics">cs.GR</span> </div> </div> <p class="title is-5 mathjax"> DART: A Diffusion-Based Autoregressive Motion Model for Real-Time Text-Driven Motion Control </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaifeng Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+G">Gen Li</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+S">Siyu Tang</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="2410.05260v1-abstract-short" style="display: inline;"> Text-conditioned human motion generation, which allows for user interaction through natural language, has become increasingly popular. Existing methods typically generate short, isolated motions based on a single input sentence. However, human motions are continuous and can extend over long periods, carrying rich semantics. Creating long, complex motions that precisely respond to streams of text d… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.05260v1-abstract-full').style.display = 'inline'; document.getElementById('2410.05260v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.05260v1-abstract-full" style="display: none;"> Text-conditioned human motion generation, which allows for user interaction through natural language, has become increasingly popular. Existing methods typically generate short, isolated motions based on a single input sentence. However, human motions are continuous and can extend over long periods, carrying rich semantics. Creating long, complex motions that precisely respond to streams of text descriptions, particularly in an online and real-time setting, remains a significant challenge. Furthermore, incorporating spatial constraints into text-conditioned motion generation presents additional challenges, as it requires aligning the motion semantics specified by text descriptions with geometric information, such as goal locations and 3D scene geometry. To address these limitations, we propose DART, a Diffusion-based Autoregressive motion primitive model for Real-time Text-driven motion control. Our model, DART, effectively learns a compact motion primitive space jointly conditioned on motion history and text inputs using latent diffusion models. By autoregressively generating motion primitives based on the preceding history and current text input, DART enables real-time, sequential motion generation driven by natural language descriptions. Additionally, the learned motion primitive space allows for precise spatial motion control, which we formulate either as a latent noise optimization problem or as a Markov decision process addressed through reinforcement learning. We present effective algorithms for both approaches, demonstrating our model's versatility and superior performance in various motion synthesis tasks. Experiments show our method outperforms existing baselines in motion realism, efficiency, and controllability. Video results are available on the project page: https://zkf1997.github.io/DART/. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.05260v1-abstract-full').style.display = 'none'; document.getElementById('2410.05260v1-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> 7 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.04939">arXiv:2410.04939</a> <span> [<a href="https://arxiv.org/pdf/2410.04939">pdf</a>, <a href="https://arxiv.org/format/2410.04939">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> PRFusion: Toward Effective and Robust Multi-Modal Place Recognition with Image and Point Cloud Fusion </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+S">Sijie Wang</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+Q">Qiyu Kang</a>, <a href="/search/cs?searchtype=author&query=She%2C+R">Rui She</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Song%2C+Y">Yang Song</a>, <a href="/search/cs?searchtype=author&query=Tay%2C+W+P">Wee Peng Tay</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="2410.04939v1-abstract-short" style="display: inline;"> Place recognition plays a crucial role in the fields of robotics and computer vision, finding applications in areas such as autonomous driving, mapping, and localization. Place recognition identifies a place using query sensor data and a known database. One of the main challenges is to develop a model that can deliver accurate results while being robust to environmental variations. We propose two… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.04939v1-abstract-full').style.display = 'inline'; document.getElementById('2410.04939v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.04939v1-abstract-full" style="display: none;"> Place recognition plays a crucial role in the fields of robotics and computer vision, finding applications in areas such as autonomous driving, mapping, and localization. Place recognition identifies a place using query sensor data and a known database. One of the main challenges is to develop a model that can deliver accurate results while being robust to environmental variations. We propose two multi-modal place recognition models, namely PRFusion and PRFusion++. PRFusion utilizes global fusion with manifold metric attention, enabling effective interaction between features without requiring camera-LiDAR extrinsic calibrations. In contrast, PRFusion++ assumes the availability of extrinsic calibrations and leverages pixel-point correspondences to enhance feature learning on local windows. Additionally, both models incorporate neural diffusion layers, which enable reliable operation even in challenging environments. We verify the state-of-the-art performance of both models on three large-scale benchmarks. Notably, they outperform existing models by a substantial margin of +3.0 AR@1 on the demanding Boreas dataset. Furthermore, we conduct ablation studies to validate the effectiveness of our proposed methods. The codes are available at: https://github.com/sijieaaa/PRFusion <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.04939v1-abstract-full').style.display = 'none'; document.getElementById('2410.04939v1-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> 7 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">accepted by IEEE TITS 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.03038">arXiv:2410.03038</a> <span> [<a href="https://arxiv.org/pdf/2410.03038">pdf</a>, <a href="https://arxiv.org/format/2410.03038">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="Computer Vision and Pattern Recognition">cs.CV</span> </div> <div class="is-inline-block" style="margin-left: 0.5rem"> <div class="tags has-addons"> <span class="tag is-dark is-size-7">doi</span> <span class="tag is-light is-size-7"><a class="" href="https://doi.org/10.1145/3627673.3680045">10.1145/3627673.3680045 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> CPFD: Confidence-aware Privileged Feature Distillation for Short Video Classification </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shi%2C+J">Jinghao Shi</a>, <a href="/search/cs?searchtype=author&query=Shen%2C+X">Xiang Shen</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaili Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xuedong Wang</a>, <a href="/search/cs?searchtype=author&query=Wen%2C+V">Vera Wen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zixuan Wang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+Y">Yifan Wu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhixin Zhang</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="2410.03038v2-abstract-short" style="display: inline;"> Dense features, customized for different business scenarios, are essential in short video classification. However, their complexity, specific adaptation requirements, and high computational costs make them resource-intensive and less accessible during online inference. Consequently, these dense features are categorized as `Privileged Dense Features'.Meanwhile, end-to-end multi-modal models have sh… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.03038v2-abstract-full').style.display = 'inline'; document.getElementById('2410.03038v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.03038v2-abstract-full" style="display: none;"> Dense features, customized for different business scenarios, are essential in short video classification. However, their complexity, specific adaptation requirements, and high computational costs make them resource-intensive and less accessible during online inference. Consequently, these dense features are categorized as `Privileged Dense Features'.Meanwhile, end-to-end multi-modal models have shown promising results in numerous computer vision tasks. In industrial applications, prioritizing end-to-end multi-modal features, can enhance efficiency but often leads to the loss of valuable information from historical privileged dense features. To integrate both features while maintaining efficiency and manageable resource costs, we present Confidence-aware Privileged Feature Distillation (CPFD), which empowers features of an end-to-end multi-modal model by adaptively distilling privileged features during training. Unlike existing privileged feature distillation (PFD) methods, which apply uniform weights to all instances during distillation, potentially causing unstable performance across different business scenarios and a notable performance gap between teacher model (Dense Feature enhanced multimodal-model DF-X-VLM) and student model (multimodal-model only X-VLM), our CPFD leverages confidence scores derived from the teacher model to adaptively mitigate the performance variance with the student model. We conducted extensive offline experiments on five diverse tasks demonstrating that CPFD improves the video classification F1 score by 6.76% compared with end-to-end multimodal-model (X-VLM) and by 2.31% with vanilla PFD on-average. And it reduces the performance gap by 84.6% and achieves results comparable to teacher model DF-X-VLM. The effectiveness of CPFD is further substantiated by online experiments, and our framework has been deployed in production systems for over a dozen models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.03038v2-abstract-full').style.display = 'none'; document.getElementById('2410.03038v2-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 3 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 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">Camera ready for CIKM 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.20500">arXiv:2409.20500</a> <span> [<a href="https://arxiv.org/pdf/2409.20500">pdf</a>, <a href="https://arxiv.org/format/2409.20500">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Multimedia">cs.MM</span> </div> </div> <p class="title is-5 mathjax"> FreeMask: Rethinking the Importance of Attention Masks for Zero-Shot Video Editing </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Cai%2C+L">Lingling Cai</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+H">Hangjie Yuan</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yingya Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+S">Shiwei Zhang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+K">Kejie Huang</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="2409.20500v1-abstract-short" style="display: inline;"> Text-to-video diffusion models have made remarkable advancements. Driven by their ability to generate temporally coherent videos, research on zero-shot video editing using these fundamental models has expanded rapidly. To enhance editing quality, structural controls are frequently employed in video editing. Among these techniques, cross-attention mask control stands out for its effectiveness and e… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.20500v1-abstract-full').style.display = 'inline'; document.getElementById('2409.20500v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.20500v1-abstract-full" style="display: none;"> Text-to-video diffusion models have made remarkable advancements. Driven by their ability to generate temporally coherent videos, research on zero-shot video editing using these fundamental models has expanded rapidly. To enhance editing quality, structural controls are frequently employed in video editing. Among these techniques, cross-attention mask control stands out for its effectiveness and efficiency. However, when cross-attention masks are naively applied to video editing, they can introduce artifacts such as blurring and flickering. Our experiments uncover a critical factor overlooked in previous video editing research: cross-attention masks are not consistently clear but vary with model structure and denoising timestep. To address this issue, we propose the metric Mask Matching Cost (MMC) that quantifies this variability and propose FreeMask, a method for selecting optimal masks tailored to specific video editing tasks. Using MMC-selected masks, we further improve the masked fusion mechanism within comprehensive attention features, e.g., temp, cross, and self-attention modules. Our approach can be seamlessly integrated into existing zero-shot video editing frameworks with better performance, requiring no control assistance or parameter fine-tuning but enabling adaptive decoupling of unedited semantic layouts with mask precision control. Extensive experiments demonstrate that FreeMask achieves superior semantic fidelity, temporal consistency, and editing quality compared to state-of-the-art methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.20500v1-abstract-full').style.display = 'none'; document.getElementById('2409.20500v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 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">Video Editing</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.19130">arXiv:2409.19130</a> <span> [<a href="https://arxiv.org/pdf/2409.19130">pdf</a>, <a href="https://arxiv.org/format/2409.19130">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Image and Video Processing">eess.IV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</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"> Multi-modal Cross-domain Self-supervised Pre-training for fMRI and EEG Fusion </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wei%2C+X">Xinxu Wei</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kanhao Zhao</a>, <a href="/search/cs?searchtype=author&query=Jiao%2C+Y">Yong Jiao</a>, <a href="/search/cs?searchtype=author&query=Carlisle%2C+N+B">Nancy B. Carlisle</a>, <a href="/search/cs?searchtype=author&query=Xie%2C+H">Hua Xie</a>, <a href="/search/cs?searchtype=author&query=Fonzo%2C+G+A">Gregory A. Fonzo</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yu Zhang</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="2409.19130v1-abstract-short" style="display: inline;"> Neuroimaging techniques including functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) have shown promise in detecting functional abnormalities in various brain disorders. However, existing studies often focus on a single domain or modality, neglecting the valuable complementary information offered by multiple domains from both fMRI and EEG, which is crucial for a comprehens… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.19130v1-abstract-full').style.display = 'inline'; document.getElementById('2409.19130v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.19130v1-abstract-full" style="display: none;"> Neuroimaging techniques including functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) have shown promise in detecting functional abnormalities in various brain disorders. However, existing studies often focus on a single domain or modality, neglecting the valuable complementary information offered by multiple domains from both fMRI and EEG, which is crucial for a comprehensive representation of disorder pathology. This limitation poses a challenge in effectively leveraging the synergistic information derived from these modalities. To address this, we propose a Multi-modal Cross-domain Self-supervised Pre-training Model (MCSP), a novel approach that leverages self-supervised learning to synergize multi-modal information across spatial, temporal, and spectral domains. Our model employs cross-domain self-supervised loss that bridges domain differences by implementing domain-specific data augmentation and contrastive loss, enhancing feature discrimination. Furthermore, MCSP introduces cross-modal self-supervised loss to capitalize on the complementary information of fMRI and EEG, facilitating knowledge distillation within domains and maximizing cross-modal feature convergence. We constructed a large-scale pre-training dataset and pretrained MCSP model by leveraging proposed self-supervised paradigms to fully harness multimodal neuroimaging data. Through comprehensive experiments, we have demonstrated the superior performance and generalizability of our model on multiple classification tasks. Our study contributes a significant advancement in the fusion of fMRI and EEG, marking a novel integration of cross-domain features, which enriches the existing landscape of neuroimaging research, particularly within the context of mental disorder studies. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.19130v1-abstract-full').style.display = 'none'; document.getElementById('2409.19130v1-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> 27 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.17790">arXiv:2409.17790</a> <span> [<a href="https://arxiv.org/pdf/2409.17790">pdf</a>, <a href="https://arxiv.org/format/2409.17790">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="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> CASPFormer: Trajectory Prediction from BEV Images with Deformable Attention </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yadav%2C+H">Harsh Yadav</a>, <a href="/search/cs?searchtype=author&query=Schaefer%2C+M">Maximilian Schaefer</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Meisen%2C+T">Tobias Meisen</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="2409.17790v1-abstract-short" style="display: inline;"> Motion prediction is an important aspect for Autonomous Driving (AD) and Advance Driver Assistance Systems (ADAS). Current state-of-the-art motion prediction methods rely on High Definition (HD) maps for capturing the surrounding context of the ego vehicle. Such systems lack scalability in real-world deployment as HD maps are expensive to produce and update in real-time. To overcome this issue, we… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17790v1-abstract-full').style.display = 'inline'; document.getElementById('2409.17790v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.17790v1-abstract-full" style="display: none;"> Motion prediction is an important aspect for Autonomous Driving (AD) and Advance Driver Assistance Systems (ADAS). Current state-of-the-art motion prediction methods rely on High Definition (HD) maps for capturing the surrounding context of the ego vehicle. Such systems lack scalability in real-world deployment as HD maps are expensive to produce and update in real-time. To overcome this issue, we propose Context Aware Scene Prediction Transformer (CASPFormer), which can perform multi-modal motion prediction from rasterized Bird-Eye-View (BEV) images. Our system can be integrated with any upstream perception module that is capable of generating BEV images. Moreover, CASPFormer directly decodes vectorized trajectories without any postprocessing. Trajectories are decoded recurrently using deformable attention, as it is computationally efficient and provides the network with the ability to focus its attention on the important spatial locations of the BEV images. In addition, we also address the issue of mode collapse for generating multiple scene-consistent trajectories by incorporating learnable mode queries. We evaluate our model on the nuScenes dataset and show that it reaches state-of-the-art across multiple metrics <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17790v1-abstract-full').style.display = 'none'; document.getElementById('2409.17790v1-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> 26 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 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">Under Review at ICPR 2024, Kolkata</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.17527">arXiv:2409.17527</a> <span> [<a href="https://arxiv.org/pdf/2409.17527">pdf</a>, <a href="https://arxiv.org/format/2409.17527">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"> Data Proportion Detection for Optimized Data Management for Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Liang%2C+H">Hao Liang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Keshi Zhao</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Y">Yajie Yang</a>, <a href="/search/cs?searchtype=author&query=Cui%2C+B">Bin Cui</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+G">Guosheng Dong</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Z">Zenan Zhou</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wentao Zhang</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="2409.17527v1-abstract-short" style="display: inline;"> Large language models (LLMs) have demonstrated exceptional performance across a wide range of tasks and domains, with data preparation playing a critical role in achieving these results. Pre-training data typically combines information from multiple domains. To maximize performance when integrating data from various domains, determining the optimal data proportion is essential. However, state-of-t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17527v1-abstract-full').style.display = 'inline'; document.getElementById('2409.17527v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.17527v1-abstract-full" style="display: none;"> Large language models (LLMs) have demonstrated exceptional performance across a wide range of tasks and domains, with data preparation playing a critical role in achieving these results. Pre-training data typically combines information from multiple domains. To maximize performance when integrating data from various domains, determining the optimal data proportion is essential. However, state-of-the-art (SOTA) LLMs rarely disclose details about their pre-training data, making it difficult for researchers to identify ideal data proportions. In this paper, we introduce a new topic, \textit{data proportion detection}, which enables the automatic estimation of pre-training data proportions by analyzing the generated outputs of LLMs. We provide rigorous theoretical proofs, practical algorithms, and preliminary experimental results for data proportion detection. Based on these findings, we offer valuable insights into the challenges and future directions for effective data proportion detection and data management. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17527v1-abstract-full').style.display = 'none'; document.getElementById('2409.17527v1-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> 26 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.12454">arXiv:2409.12454</a> <span> [<a href="https://arxiv.org/pdf/2409.12454">pdf</a>, <a href="https://arxiv.org/format/2409.12454">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Signal Processing">eess.SP</span> </div> </div> <p class="title is-5 mathjax"> FoME: A Foundation Model for EEG using Adaptive Temporal-Lateral Attention Scaling </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shi%2C+E">Enze Shi</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kui Zhao</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+Q">Qilong Yuan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Jiaqi Wang</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+H">Huawen Hu</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+S">Sigang Yu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+S">Shu Zhang</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="2409.12454v1-abstract-short" style="display: inline;"> Electroencephalography (EEG) is a vital tool to measure and record brain activity in neuroscience and clinical applications, yet its potential is constrained by signal heterogeneity, low signal-to-noise ratios, and limited labeled datasets. In this paper, we propose FoME (Foundation Model for EEG), a novel approach using adaptive temporal-lateral attention scaling to address above-mentioned challe… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.12454v1-abstract-full').style.display = 'inline'; document.getElementById('2409.12454v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.12454v1-abstract-full" style="display: none;"> Electroencephalography (EEG) is a vital tool to measure and record brain activity in neuroscience and clinical applications, yet its potential is constrained by signal heterogeneity, low signal-to-noise ratios, and limited labeled datasets. In this paper, we propose FoME (Foundation Model for EEG), a novel approach using adaptive temporal-lateral attention scaling to address above-mentioned challenges. FoME is pre-trained on a diverse 1.7TB dataset of scalp and intracranial EEG recordings, comprising 745M parameters trained for 1,096k steps. Our model introduces two key innovations: a time-frequency fusion embedding technique and an adaptive time-lateral attention scaling (ATLAS) mechanism. These components synergistically capture complex temporal and spectral EEG dynamics, enabling FoME to adapt to varying patterns across diverse data streams and facilitate robust multi-channel modeling. Evaluations across four downstream tasks demonstrate FoME's superior performance in classification and forecasting applications, consistently achieving state-of-the-art results. To conclude, FoME establishes a new paradigm for EEG analysis, offering a versatile foundation that advances brain-computer interfaces, clinical diagnostics, and cognitive research across neuroscience and related fields. Our code will be available at https://github.com/1061413241/FoME. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.12454v1-abstract-full').style.display = 'none'; document.getElementById('2409.12454v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.03029">arXiv:2409.03029</a> <span> [<a href="https://arxiv.org/pdf/2409.03029">pdf</a>, <a href="https://arxiv.org/format/2409.03029">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> GreenWhisk: Emission-Aware Computing for Serverless Platform </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Serenari%2C+J">Jayden Serenari</a>, <a href="/search/cs?searchtype=author&query=Sreekumar%2C+S">Sreekanth Sreekumar</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiwen Zhao</a>, <a href="/search/cs?searchtype=author&query=Sarkar%2C+S">Saurabh Sarkar</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+S">Stephen Lee</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="2409.03029v1-abstract-short" style="display: inline;"> Serverless computing is an emerging cloud computing abstraction wherein the cloud platform transparently manages all resources, including explicitly provisioning resources and geographical load balancing when the demand for service spikes. Users provide code as functions, and the cloud platform runs these functions handling all aspects of function execution. While prior work has primarily focused… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.03029v1-abstract-full').style.display = 'inline'; document.getElementById('2409.03029v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.03029v1-abstract-full" style="display: none;"> Serverless computing is an emerging cloud computing abstraction wherein the cloud platform transparently manages all resources, including explicitly provisioning resources and geographical load balancing when the demand for service spikes. Users provide code as functions, and the cloud platform runs these functions handling all aspects of function execution. While prior work has primarily focused on optimizing performance, this paper focuses on reducing the carbon footprint of these systems making variations in grid carbon intensity and intermittency from renewables transparent to the user. We introduce GreenWhisk, a carbon-aware serverless computing platform built upon Apache OpenWhisk, operating in two modes - grid-connected and grid-isolated - addressing intermittency challenges arising from renewables and the grid's carbon footprint. Moreover, we develop carbon-aware load balancing algorithms that leverage energy and carbon information to reduce the carbon footprint. Our evaluation results show that GreenWhisk can easily incorporate carbon-aware algorithms, thereby reducing the carbon footprint of functions without significantly impacting the performance of function execution. In doing so, our system design enables the integration of new carbon-aware strategies into a serverless computing platform. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.03029v1-abstract-full').style.display = 'none'; document.getElementById('2409.03029v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 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">11 pages, 13 figures, IC2E 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.01555">arXiv:2409.01555</a> <span> [<a href="https://arxiv.org/pdf/2409.01555">pdf</a>, <a href="https://arxiv.org/format/2409.01555">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</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"> EA-RAS: Towards Efficient and Accurate End-to-End Reconstruction of Anatomical Skeleton </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Peng%2C+Z">Zhiheng Peng</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xiaoran Chen</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+L">Li Ma</a>, <a href="/search/cs?searchtype=author&query=Xia%2C+S">Siyu Xia</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+C">Changjie Fan</a>, <a href="/search/cs?searchtype=author&query=Shang%2C+W">Weijian Shang</a>, <a href="/search/cs?searchtype=author&query=Jing%2C+W">Wei Jing</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="2409.01555v1-abstract-short" style="display: inline;"> Efficient, accurate and low-cost estimation of human skeletal information is crucial for a range of applications such as biology education and human-computer interaction. However, current simple skeleton models, which are typically based on 2D-3D joint points, fall short in terms of anatomical fidelity, restricting their utility in fields. On the other hand, more complex models while anatomically… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.01555v1-abstract-full').style.display = 'inline'; document.getElementById('2409.01555v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.01555v1-abstract-full" style="display: none;"> Efficient, accurate and low-cost estimation of human skeletal information is crucial for a range of applications such as biology education and human-computer interaction. However, current simple skeleton models, which are typically based on 2D-3D joint points, fall short in terms of anatomical fidelity, restricting their utility in fields. On the other hand, more complex models while anatomically precise, are hindered by sophisticate multi-stage processing and the need for extra data like skin meshes, making them unsuitable for real-time applications. To this end, we propose the EA-RAS (Towards Efficient and Accurate End-to-End Reconstruction of Anatomical Skeleton), a single-stage, lightweight, and plug-and-play anatomical skeleton estimator that can provide real-time, accurate anatomically realistic skeletons with arbitrary pose using only a single RGB image input. Additionally, EA-RAS estimates the conventional human-mesh model explicitly, which not only enhances the functionality but also leverages the outside skin information by integrating features into the inside skeleton modeling process. In this work, we also develop a progressive training strategy and integrated it with an enhanced optimization process, enabling the network to obtain initial weights using only a small skin dataset and achieve self-supervision in skeleton reconstruction. Besides, we also provide an optional lightweight post-processing optimization strategy to further improve accuracy for scenarios that prioritize precision over real-time processing. The experiments demonstrated that our regression method is over 800 times faster than existing methods, meeting real-time requirements. Additionally, the post-processing optimization strategy provided can enhance reconstruction accuracy by over 50% and achieve a speed increase of more than 7 times. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.01555v1-abstract-full').style.display = 'none'; document.getElementById('2409.01555v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 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">13 pages,15 figures</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.00997">arXiv:2409.00997</a> <span> [<a href="https://arxiv.org/pdf/2409.00997">pdf</a>, <a href="https://arxiv.org/format/2409.00997">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"> DataSculpt: Crafting Data Landscapes for Long-Context LLMs through Multi-Objective Partitioning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lu%2C+K">Keer Lu</a>, <a href="/search/cs?searchtype=author&query=Nie%2C+X">Xiaonan Nie</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+Z">Zheng Liang</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+D">Da Pan</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+S">Shusen Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Keshi Zhao</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+W">Weipeng Chen</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Z">Zenan Zhou</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+G">Guosheng Dong</a>, <a href="/search/cs?searchtype=author&query=Cui%2C+B">Bin Cui</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wentao Zhang</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="2409.00997v2-abstract-short" style="display: inline;"> In recent years, Large Language Models (LLMs) have demonstrated significant improvements across a variety of tasks, one of which is the long-context capability. The key to improving long-context performance lies in effective data organization and management strategies that integrate data from multiple domains and optimize the context window during training. Through extensive experimental analysis,… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.00997v2-abstract-full').style.display = 'inline'; document.getElementById('2409.00997v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.00997v2-abstract-full" style="display: none;"> In recent years, Large Language Models (LLMs) have demonstrated significant improvements across a variety of tasks, one of which is the long-context capability. The key to improving long-context performance lies in effective data organization and management strategies that integrate data from multiple domains and optimize the context window during training. Through extensive experimental analysis, we identified three key challenges in designing effective data management strategies that enable the model to achieve long-context capability without sacrificing performance in other tasks: (1) a shortage of long documents across multiple domains, (2) effective construction of context windows, and (3) efficient organization of large-scale datasets. To address these challenges, we introduce DataSculpt, a novel data management framework designed for long-context training. We first formulate the organization of training data as a multi-objective combinatorial optimization problem, focusing on attributes including relevance, homogeneity, integrity, and efficiency. Specifically, our approach utilizes a coarse-to-fine methodology to optimize training data organization both efficiently and effectively. We begin by clustering the data based on semantic similarity (coarse), followed by a multi-objective greedy search within each cluster to score and concatenate documents into various context windows (fine). Our comprehensive evaluations demonstrate that DataSculpt significantly enhances long-context training performance, resulting in improvements of 18.09% in retrieval augmentation, 21.23% in summarization, 21.27% in reading comprehension, and a 3.81% increase in code completion, while also maintaining overall model proficiency with a 4.88% improvement. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.00997v2-abstract-full').style.display = 'none'; document.getElementById('2409.00997v2-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 2 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.14267">arXiv:2408.14267</a> <span> [<a href="https://arxiv.org/pdf/2408.14267">pdf</a>, <a href="https://arxiv.org/format/2408.14267">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="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> 1-Bit FQT: Pushing the Limit of Fully Quantized Training to 1-bit </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Gao%2C+C">Chang Gao</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+J">Jianfei Chen</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Jiaqi Wang</a>, <a href="/search/cs?searchtype=author&query=Jing%2C+L">Liping Jing</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="2408.14267v1-abstract-short" style="display: inline;"> Fully quantized training (FQT) accelerates the training of deep neural networks by quantizing the activations, weights, and gradients into lower precision. To explore the ultimate limit of FQT (the lowest achievable precision), we make a first attempt to 1-bit FQT. We provide a theoretical analysis of FQT based on Adam and SGD, revealing that the gradient variance influences the convergence of FQT… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.14267v1-abstract-full').style.display = 'inline'; document.getElementById('2408.14267v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.14267v1-abstract-full" style="display: none;"> Fully quantized training (FQT) accelerates the training of deep neural networks by quantizing the activations, weights, and gradients into lower precision. To explore the ultimate limit of FQT (the lowest achievable precision), we make a first attempt to 1-bit FQT. We provide a theoretical analysis of FQT based on Adam and SGD, revealing that the gradient variance influences the convergence of FQT. Building on these theoretical results, we introduce an Activation Gradient Pruning (AGP) strategy. The strategy leverages the heterogeneity of gradients by pruning less informative gradients and enhancing the numerical precision of remaining gradients to mitigate gradient variance. Additionally, we propose Sample Channel joint Quantization (SCQ), which utilizes different quantization strategies in the computation of weight gradients and activation gradients to ensure that the method is friendly to low-bitwidth hardware. Finally, we present a framework to deploy our algorithm. For fine-tuning VGGNet-16 and ResNet-18 on multiple datasets, our algorithm achieves an average accuracy improvement of approximately 6%, compared to per-sample quantization. Moreover, our training speedup can reach a maximum of 5.13x compared to full precision training. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.14267v1-abstract-full').style.display = 'none'; document.getElementById('2408.14267v1-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> 26 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.12133">arXiv:2408.12133</a> <span> [<a href="https://arxiv.org/pdf/2408.12133">pdf</a>, <a href="https://arxiv.org/format/2408.12133">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Self-supervised Learning for Geospatial AI: A Survey </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yile Chen</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+W">Weiming Huang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiqi Zhao</a>, <a href="/search/cs?searchtype=author&query=Jiang%2C+Y">Yue Jiang</a>, <a href="/search/cs?searchtype=author&query=Cong%2C+G">Gao Cong</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="2408.12133v1-abstract-short" style="display: inline;"> The proliferation of geospatial data in urban and territorial environments has significantly facilitated the development of geospatial artificial intelligence (GeoAI) across various urban applications. Given the vast yet inherently sparse labeled nature of geospatial data, there is a critical need for techniques that can effectively leverage such data without heavy reliance on labeled datasets. Th… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.12133v1-abstract-full').style.display = 'inline'; document.getElementById('2408.12133v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.12133v1-abstract-full" style="display: none;"> The proliferation of geospatial data in urban and territorial environments has significantly facilitated the development of geospatial artificial intelligence (GeoAI) across various urban applications. Given the vast yet inherently sparse labeled nature of geospatial data, there is a critical need for techniques that can effectively leverage such data without heavy reliance on labeled datasets. This requirement aligns with the principles of self-supervised learning (SSL), which has attracted increasing attention for its adoption in geospatial data. This paper conducts a comprehensive and up-to-date survey of SSL techniques applied to or developed for three primary data (geometric) types prevalent in geospatial vector data: points, polylines, and polygons. We systematically categorize various SSL techniques into predictive and contrastive methods, discussing their application with respect to each data type in enhancing generalization across various downstream tasks. Furthermore, we review the emerging trends of SSL for GeoAI, and several task-specific SSL techniques. Finally, we discuss several key challenges in the current research and outline promising directions for future investigation. By presenting a structured analysis of relevant studies, this paper aims to inspire continued advancements in the integration of SSL with GeoAI, encouraging innovative methods to harnessing the power of geospatial data. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.12133v1-abstract-full').style.display = 'none'; document.getElementById('2408.12133v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.11971">arXiv:2408.11971</a> <span> [<a href="https://arxiv.org/pdf/2408.11971">pdf</a>, <a href="https://arxiv.org/format/2408.11971">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> HoSZp: An Efficient Homomorphic Error-bounded Lossy Compressor for Scientific Data </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Agarwal%2C+T">Tripti Agarwal</a>, <a href="/search/cs?searchtype=author&query=Di%2C+S">Sheng Di</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+J">Jiajun Huang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yafan Huang</a>, <a href="/search/cs?searchtype=author&query=Gopalakrishnan%2C+G">Ganesh Gopalakrishnan</a>, <a href="/search/cs?searchtype=author&query=Underwood%2C+R">Robert Underwood</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+X">Xin Liang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+G">Guanpeng Li</a>, <a href="/search/cs?searchtype=author&query=Cappello%2C+F">Franck Cappello</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="2408.11971v1-abstract-short" style="display: inline;"> Error-bounded lossy compression has been a critical technique to significantly reduce the sheer amounts of simulation datasets for high-performance computing (HPC) scientific applications while effectively controlling the data distortion based on user-specified error bound. In many real-world use cases, users must perform computational operations on the compressed data (a.k.a. homomorphic compress… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.11971v1-abstract-full').style.display = 'inline'; document.getElementById('2408.11971v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.11971v1-abstract-full" style="display: none;"> Error-bounded lossy compression has been a critical technique to significantly reduce the sheer amounts of simulation datasets for high-performance computing (HPC) scientific applications while effectively controlling the data distortion based on user-specified error bound. In many real-world use cases, users must perform computational operations on the compressed data (a.k.a. homomorphic compression). However, none of the existing error-bounded lossy compressors support the homomorphism, inevitably resulting in undesired decompression costs. In this paper, we propose a novel homomorphic error-bounded lossy compressor (called HoSZp), which supports not only error-bounding features but efficient computations (including negation, addition, multiplication, mean, variance, etc.) on the compressed data without the complete decompression step, which is the first attempt to the best of our knowledge. We develop several optimization strategies to maximize the overall compression ratio and execution performance. We evaluate HoSZp compared to other state-of-the-art lossy compressors based on multiple real-world scientific application datasets. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.11971v1-abstract-full').style.display = 'none'; document.getElementById('2408.11971v1-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 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">12 pages, 7 figures, 9 tables</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.09347">arXiv:2408.09347</a> <span> [<a href="https://arxiv.org/pdf/2408.09347">pdf</a>, <a href="https://arxiv.org/format/2408.09347">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> S^3D-NeRF: Single-Shot Speech-Driven Neural Radiance Field for High Fidelity Talking Head Synthesis </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+D">Dongze Li</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+W">Wei Wang</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+Y">Yifeng Ma</a>, <a href="/search/cs?searchtype=author&query=Peng%2C+B">Bo Peng</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yingya Zhang</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+J">Jing Dong</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="2408.09347v1-abstract-short" style="display: inline;"> Talking head synthesis is a practical technique with wide applications. Current Neural Radiance Field (NeRF) based approaches have shown their superiority on driving one-shot talking heads with videos or signals regressed from audio. However, most of them failed to take the audio as driven information directly, unable to enjoy the flexibility and availability of speech. Since mapping audio signals… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.09347v1-abstract-full').style.display = 'inline'; document.getElementById('2408.09347v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.09347v1-abstract-full" style="display: none;"> Talking head synthesis is a practical technique with wide applications. Current Neural Radiance Field (NeRF) based approaches have shown their superiority on driving one-shot talking heads with videos or signals regressed from audio. However, most of them failed to take the audio as driven information directly, unable to enjoy the flexibility and availability of speech. Since mapping audio signals to face deformation is non-trivial, we design a Single-Shot Speech-Driven Neural Radiance Field (S^3D-NeRF) method in this paper to tackle the following three difficulties: learning a representative appearance feature for each identity, modeling motion of different face regions with audio, and keeping the temporal consistency of the lip area. To this end, we introduce a Hierarchical Facial Appearance Encoder to learn multi-scale representations for catching the appearance of different speakers, and elaborate a Cross-modal Facial Deformation Field to perform speech animation according to the relationship between the audio signal and different face regions. Moreover, to enhance the temporal consistency of the important lip area, we introduce a lip-sync discriminator to penalize the out-of-sync audio-visual sequences. Extensive experiments have shown that our S^3D-NeRF surpasses previous arts on both video fidelity and audio-lip synchronization. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.09347v1-abstract-full').style.display = 'none'; document.getElementById('2408.09347v1-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> 17 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">ECCV 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.06840">arXiv:2408.06840</a> <span> [<a href="https://arxiv.org/pdf/2408.06840">pdf</a>, <a href="https://arxiv.org/format/2408.06840">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> Dynamic and Compressive Adaptation of Transformers From Images to Videos </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+G">Guozhen Zhang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jingyu Liu</a>, <a href="/search/cs?searchtype=author&query=Cao%2C+S">Shengming Cao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+X">Xiaotong Zhao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kevin Zhao</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+K">Kai Ma</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+L">Limin Wang</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="2408.06840v2-abstract-short" style="display: inline;"> Recently, the remarkable success of pre-trained Vision Transformers (ViTs) from image-text matching has sparked an interest in image-to-video adaptation. However, most current approaches retain the full forward pass for each frame, leading to a high computation overhead for processing entire videos. In this paper, we present InTI, a novel approach for compressive image-to-video adaptation using dy… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.06840v2-abstract-full').style.display = 'inline'; document.getElementById('2408.06840v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.06840v2-abstract-full" style="display: none;"> Recently, the remarkable success of pre-trained Vision Transformers (ViTs) from image-text matching has sparked an interest in image-to-video adaptation. However, most current approaches retain the full forward pass for each frame, leading to a high computation overhead for processing entire videos. In this paper, we present InTI, a novel approach for compressive image-to-video adaptation using dynamic Inter-frame Token Interpolation. InTI aims to softly preserve the informative tokens without disrupting their coherent spatiotemporal structure. Specifically, each token pair at identical positions within neighbor frames is linearly aggregated into a new token, where the aggregation weights are generated by a multi-scale context-aware network. In this way, the information of neighbor frames can be adaptively compressed in a point-by-point manner, thereby effectively reducing the number of processed frames by half each time. Importantly, InTI can be seamlessly integrated with existing adaptation methods, achieving strong performance without extra-complex design. On Kinetics-400, InTI reaches a top-1 accuracy of 87.1 with a remarkable 37.5% reduction in GFLOPs compared to naive adaptation. When combined with additional temporal modules, InTI achieves a top-1 accuracy of 87.6 with a 37% reduction in GFLOPs. Similar conclusions have been verified in other common datasets. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.06840v2-abstract-full').style.display = 'none'; document.getElementById('2408.06840v2-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> 13 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 13 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.19349">arXiv:2407.19349</a> <span> [<a href="https://arxiv.org/pdf/2407.19349">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Quantitative Methods">q-bio.QM</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"> Predicting T-Cell Receptor Specificity </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tu%2C+T">Tengyao Tu</a>, <a href="/search/cs?searchtype=author&query=Zeng%2C+W">Wei Zeng</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhenyu Zhang</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="2407.19349v1-abstract-short" style="display: inline;"> Researching the specificity of TCR contributes to the development of immunotherapy and provides new opportunities and strategies for personalized cancer immunotherapy. Therefore, we established a TCR generative specificity detection framework consisting of an antigen selector and a TCR classifier based on the Random Forest algorithm, aiming to efficiently screen out TCRs and target antigens and ac… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.19349v1-abstract-full').style.display = 'inline'; document.getElementById('2407.19349v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.19349v1-abstract-full" style="display: none;"> Researching the specificity of TCR contributes to the development of immunotherapy and provides new opportunities and strategies for personalized cancer immunotherapy. Therefore, we established a TCR generative specificity detection framework consisting of an antigen selector and a TCR classifier based on the Random Forest algorithm, aiming to efficiently screen out TCRs and target antigens and achieve TCR specificity prediction. Furthermore, we used the k-fold validation method to compare the performance of our model with ordinary deep learning methods. The result proves that adding a classifier to the model based on the random forest algorithm is very effective, and our model generally outperforms ordinary deep learning methods. Moreover, we put forward feasible optimization suggestions for the shortcomings and challenges of our model found during model implementation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.19349v1-abstract-full').style.display = 'none'; document.getElementById('2407.19349v1-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> 27 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.14326">arXiv:2407.14326</a> <span> [<a href="https://arxiv.org/pdf/2407.14326">pdf</a>, <a href="https://arxiv.org/ps/2407.14326">ps</a>, <a href="https://arxiv.org/format/2407.14326">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</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"> Panoptic Segmentation of Mammograms with Text-To-Image Diffusion Model </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Prokop%2C+J">Jakub Prokop</a>, <a href="/search/cs?searchtype=author&query=Tordera%2C+J+M">Javier Montalt Tordera</a>, <a href="/search/cs?searchtype=author&query=Mohammadi%2C+S">Sadegh Mohammadi</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="2407.14326v1-abstract-short" style="display: inline;"> Mammography is crucial for breast cancer surveillance and early diagnosis. However, analyzing mammography images is a demanding task for radiologists, who often review hundreds of mammograms daily, leading to overdiagnosis and overtreatment. Computer-Aided Diagnosis (CAD) systems have been developed to assist in this process, but their capabilities, particularly in lesion segmentation, remained li… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.14326v1-abstract-full').style.display = 'inline'; document.getElementById('2407.14326v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.14326v1-abstract-full" style="display: none;"> Mammography is crucial for breast cancer surveillance and early diagnosis. However, analyzing mammography images is a demanding task for radiologists, who often review hundreds of mammograms daily, leading to overdiagnosis and overtreatment. Computer-Aided Diagnosis (CAD) systems have been developed to assist in this process, but their capabilities, particularly in lesion segmentation, remained limited. With the contemporary advances in deep learning their performance may be improved. Recently, vision-language diffusion models emerged, demonstrating outstanding performance in image generation and transferability to various downstream tasks. We aim to harness their capabilities for breast lesion segmentation in a panoptic setting, which encompasses both semantic and instance-level predictions. Specifically, we propose leveraging pretrained features from a Stable Diffusion model as inputs to a state-of-the-art panoptic segmentation architecture, resulting in accurate delineation of individual breast lesions. To bridge the gap between natural and medical imaging domains, we incorporated a mammography-specific MAM-E diffusion model and BiomedCLIP image and text encoders into this framework. We evaluated our approach on two recently published mammography datasets, CDD-CESM and VinDr-Mammo. For the instance segmentation task, we noted 40.25 AP0.1 and 46.82 AP0.05, as well as 25.44 PQ0.1 and 26.92 PQ0.05. For the semantic segmentation task, we achieved Dice scores of 38.86 and 40.92, respectively. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.14326v1-abstract-full').style.display = 'none'; document.getElementById('2407.14326v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 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">13 pages, 4 figures. Submitted to Deep Generative Models workshop @ MICCAI 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.09250">arXiv:2407.09250</a> <span> [<a href="https://arxiv.org/pdf/2407.09250">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Networking and Internet Architecture">cs.NI</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"> FedsLLM: Federated Split Learning for Large Language Models over Communication Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Z">Zhaohui Yang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+C">Chongwen Huang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xiaoming Chen</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhaoyang Zhang</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="2407.09250v1-abstract-short" style="display: inline;"> Addressing the challenges of deploying large language models in wireless communication networks, this paper combines low-rank adaptation technology (LoRA) with the splitfed learning framework to propose the federated split learning for large language models (FedsLLM) framework. The method introduced in this paper utilizes LoRA technology to reduce processing loads by dividing the network into clie… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.09250v1-abstract-full').style.display = 'inline'; document.getElementById('2407.09250v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.09250v1-abstract-full" style="display: none;"> Addressing the challenges of deploying large language models in wireless communication networks, this paper combines low-rank adaptation technology (LoRA) with the splitfed learning framework to propose the federated split learning for large language models (FedsLLM) framework. The method introduced in this paper utilizes LoRA technology to reduce processing loads by dividing the network into client subnetworks and server subnetworks. It leverages a federated server to aggregate and update client models. As the training data are transmitted through a wireless network between clients and both main and federated servers, the training delay is determined by the learning accuracy and the allocation of communication bandwidth. This paper models the minimization of the training delay by integrating computation and communication optimization, simplifying the optimization problem into a convex problem to find the optimal solution. Additionally, it presents a lemma that describes the precise solutions to this problem. Simulation results demonstrate that the proposed optimization algorithm reduces delays by an average of 47.63% compared to unoptimized scenarios. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.09250v1-abstract-full').style.display = 'none'; document.getElementById('2407.09250v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.04381">arXiv:2407.04381</a> <span> [<a href="https://arxiv.org/pdf/2407.04381">pdf</a>, <a href="https://arxiv.org/format/2407.04381">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</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"> Multi-Branch Auxiliary Fusion YOLO with Re-parameterization Heterogeneous Convolutional for accurate object detection </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yang%2C+Z">Zhiqiang Yang</a>, <a href="/search/cs?searchtype=author&query=Guan%2C+Q">Qiu Guan</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Keer Zhao</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+J">Jianmin Yang</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+X">Xinli Xu</a>, <a href="/search/cs?searchtype=author&query=Long%2C+H">Haixia Long</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+Y">Ying Tang</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="2407.04381v1-abstract-short" style="display: inline;"> Due to the effective performance of multi-scale feature fusion, Path Aggregation FPN (PAFPN) is widely employed in YOLO detectors. However, it cannot efficiently and adaptively integrate high-level semantic information with low-level spatial information simultaneously. We propose a new model named MAF-YOLO in this paper, which is a novel object detection framework with a versatile neck named Multi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.04381v1-abstract-full').style.display = 'inline'; document.getElementById('2407.04381v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.04381v1-abstract-full" style="display: none;"> Due to the effective performance of multi-scale feature fusion, Path Aggregation FPN (PAFPN) is widely employed in YOLO detectors. However, it cannot efficiently and adaptively integrate high-level semantic information with low-level spatial information simultaneously. We propose a new model named MAF-YOLO in this paper, which is a novel object detection framework with a versatile neck named Multi-Branch Auxiliary FPN (MAFPN). Within MAFPN, the Superficial Assisted Fusion (SAF) module is designed to combine the output of the backbone with the neck, preserving an optimal level of shallow information to facilitate subsequent learning. Meanwhile, the Advanced Assisted Fusion (AAF) module deeply embedded within the neck conveys a more diverse range of gradient information to the output layer. Furthermore, our proposed Re-parameterized Heterogeneous Efficient Layer Aggregation Network (RepHELAN) module ensures that both the overall model architecture and convolutional design embrace the utilization of heterogeneous large convolution kernels. Therefore, this guarantees the preservation of information related to small targets while simultaneously achieving the multi-scale receptive field. Finally, taking the nano version of MAF-YOLO for example, it can achieve 42.4% AP on COCO with only 3.76M learnable parameters and 10.51G FLOPs, and approximately outperforms YOLOv8n by about 5.1%. The source code of this work is available at: https://github.com/yang-0201/MAF-YOLO. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.04381v1-abstract-full').style.display = 'none'; document.getElementById('2407.04381v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.04267">arXiv:2407.04267</a> <span> [<a href="https://arxiv.org/pdf/2407.04267">pdf</a>, <a href="https://arxiv.org/format/2407.04267">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> A High-Quality Workflow for Multi-Resolution Scientific Data Reduction and Visualization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+D">Daoce Wang</a>, <a href="/search/cs?searchtype=author&query=Grosset%2C+P">Pascal Grosset</a>, <a href="/search/cs?searchtype=author&query=Pulido%2C+J">Jesus Pulido</a>, <a href="/search/cs?searchtype=author&query=Athawale%2C+T+M">Tushar M. Athawale</a>, <a href="/search/cs?searchtype=author&query=Tian%2C+J">Jiannan Tian</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Luki%C4%87%2C+Z">Zarija Luki膰</a>, <a href="/search/cs?searchtype=author&query=Huebl%2C+A">Axel Huebl</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhe Wang</a>, <a href="/search/cs?searchtype=author&query=Ahrens%2C+J">James Ahrens</a>, <a href="/search/cs?searchtype=author&query=Tao%2C+D">Dingwen Tao</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="2407.04267v5-abstract-short" style="display: inline;"> Multi-resolution methods such as Adaptive Mesh Refinement (AMR) can enhance storage efficiency for HPC applications generating vast volumes of data. However, their applicability is limited and cannot be universally deployed across all applications. Furthermore, integrating lossy compression with multi-resolution techniques to further boost storage efficiency encounters significant barriers. To thi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.04267v5-abstract-full').style.display = 'inline'; document.getElementById('2407.04267v5-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.04267v5-abstract-full" style="display: none;"> Multi-resolution methods such as Adaptive Mesh Refinement (AMR) can enhance storage efficiency for HPC applications generating vast volumes of data. However, their applicability is limited and cannot be universally deployed across all applications. Furthermore, integrating lossy compression with multi-resolution techniques to further boost storage efficiency encounters significant barriers. To this end, we introduce an innovative workflow that facilitates high-quality multi-resolution data compression for both uniform and AMR simulations. Initially, to extend the usability of multi-resolution techniques, our workflow employs a compression-oriented Region of Interest (ROI) extraction method, transforming uniform data into a multi-resolution format. Subsequently, to bridge the gap between multi-resolution techniques and lossy compressors, we optimize three distinct compressors, ensuring their optimal performance on multi-resolution data. Lastly, we incorporate an advanced uncertainty visualization method into our workflow to understand the potential impacts of lossy compression. Experimental evaluation demonstrates that our workflow achieves significant compression quality improvements. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.04267v5-abstract-full').style.display = 'none'; document.getElementById('2407.04267v5-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 5 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 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">camera-ready version for SC '24</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.01146">arXiv:2407.01146</a> <span> [<a href="https://arxiv.org/pdf/2407.01146">pdf</a>, <a href="https://arxiv.org/format/2407.01146">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Image and Video Processing">eess.IV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> Cross-Slice Attention and Evidential Critical Loss for Uncertainty-Aware Prostate Cancer Detection </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hung%2C+A+L+Y">Alex Ling Yu Hung</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+H">Haoxin Zheng</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kai Zhao</a>, <a href="/search/cs?searchtype=author&query=Pang%2C+K">Kaifeng Pang</a>, <a href="/search/cs?searchtype=author&query=Terzopoulos%2C+D">Demetri Terzopoulos</a>, <a href="/search/cs?searchtype=author&query=Sung%2C+K">Kyunghyun Sung</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="2407.01146v1-abstract-short" style="display: inline;"> Current deep learning-based models typically analyze medical images in either 2D or 3D albeit disregarding volumetric information or suffering sub-optimal performance due to the anisotropic resolution of MR data. Furthermore, providing an accurate uncertainty estimation is beneficial to clinicians, as it indicates how confident a model is about its prediction. We propose a novel 2.5D cross-slice a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.01146v1-abstract-full').style.display = 'inline'; document.getElementById('2407.01146v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.01146v1-abstract-full" style="display: none;"> Current deep learning-based models typically analyze medical images in either 2D or 3D albeit disregarding volumetric information or suffering sub-optimal performance due to the anisotropic resolution of MR data. Furthermore, providing an accurate uncertainty estimation is beneficial to clinicians, as it indicates how confident a model is about its prediction. We propose a novel 2.5D cross-slice attention model that utilizes both global and local information, along with an evidential critical loss, to perform evidential deep learning for the detection in MR images of prostate cancer, one of the most common cancers and a leading cause of cancer-related death in men. We perform extensive experiments with our model on two different datasets and achieve state-of-the-art performance in prostate cancer detection along with improved epistemic uncertainty estimation. The implementation of the model is available at https://github.com/aL3x-O-o-Hung/GLCSA_ECLoss. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.01146v1-abstract-full').style.display = 'none'; document.getElementById('2407.01146v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.20038">arXiv:2406.20038</a> <span> [<a href="https://arxiv.org/pdf/2406.20038">pdf</a>, <a href="https://arxiv.org/format/2406.20038">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"> BioMNER: A Dataset for Biomedical Method Entity Recognition </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tang%2C+C">Chen Tang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bohao Yang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Lv%2C+B">Bo Lv</a>, <a href="/search/cs?searchtype=author&query=Xiao%2C+C">Chenghao Xiao</a>, <a href="/search/cs?searchtype=author&query=Guerin%2C+F">Frank Guerin</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+C">Chenghua Lin</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="2406.20038v1-abstract-short" style="display: inline;"> Named entity recognition (NER) stands as a fundamental and pivotal task within the realm of Natural Language Processing. Particularly within the domain of Biomedical Method NER, this task presents notable challenges, stemming from the continual influx of domain-specific terminologies in scholarly literature. Current research in Biomedical Method (BioMethod) NER suffers from a scarcity of resources… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.20038v1-abstract-full').style.display = 'inline'; document.getElementById('2406.20038v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.20038v1-abstract-full" style="display: none;"> Named entity recognition (NER) stands as a fundamental and pivotal task within the realm of Natural Language Processing. Particularly within the domain of Biomedical Method NER, this task presents notable challenges, stemming from the continual influx of domain-specific terminologies in scholarly literature. Current research in Biomedical Method (BioMethod) NER suffers from a scarcity of resources, primarily attributed to the intricate nature of methodological concepts, which necessitate a profound understanding for precise delineation. In this study, we propose a novel dataset for biomedical method entity recognition, employing an automated BioMethod entity recognition and information retrieval system to assist human annotation. Furthermore, we comprehensively explore a range of conventional and contemporary open-domain NER methodologies, including the utilization of cutting-edge large-scale language models (LLMs) customised to our dataset. Our empirical findings reveal that the large parameter counts of language models surprisingly inhibit the effective assimilation of entity extraction patterns pertaining to biomedical methods. Remarkably, the approach, leveraging the modestly sized ALBERT model (only 11MB), in conjunction with conditional random fields (CRF), achieves state-of-the-art (SOTA) performance. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.20038v1-abstract-full').style.display = 'none'; document.getElementById('2406.20038v1-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> 28 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/2406.17962">arXiv:2406.17962</a> <span> [<a href="https://arxiv.org/pdf/2406.17962">pdf</a>, <a href="https://arxiv.org/format/2406.17962">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"> Crafting Customisable Characters with LLMs: Introducing SimsChat, a Persona-Driven Role-Playing Agent Framework </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bohao Yang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+D">Dong Liu</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+C">Chen Tang</a>, <a href="/search/cs?searchtype=author&query=Xiao%2C+C">Chenghao Xiao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+C">Chao Li</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+L">Lin Yuan</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+G">Guang Yang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+L">Lanxiao Huang</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+C">Chenghua Lin</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="2406.17962v3-abstract-short" style="display: inline;"> Large Language Models (LLMs) demonstrate a remarkable ability to comprehend human instructions and generate high-quality text. This capability allows LLMs to function as agents that can emulate human beings at a more sophisticated level, beyond the mere replication of basic human behaviours. However, there is a lack of exploring into leveraging LLMs to craft characters from diverse aspects. In thi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17962v3-abstract-full').style.display = 'inline'; document.getElementById('2406.17962v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.17962v3-abstract-full" style="display: none;"> Large Language Models (LLMs) demonstrate a remarkable ability to comprehend human instructions and generate high-quality text. This capability allows LLMs to function as agents that can emulate human beings at a more sophisticated level, beyond the mere replication of basic human behaviours. However, there is a lack of exploring into leveraging LLMs to craft characters from diverse aspects. In this work, we introduce the Customisable Conversation Agent Framework, which leverages LLMs to simulate real-world characters that can be freely customised according to various user preferences. This adaptable framework is beneficial for the design of customisable characters and role-playing agents aligned with human preferences. We propose the SimsConv dataset, which encompasses 68 different customised characters, 1,360 multi-turn role-playing dialogues, and a total of 13,971 interaction dialogues. The characters are created from several real-world elements, such as career, aspiration, trait, and skill. Building upon these foundations, we present SimsChat, a freely customisable role-playing agent. It incorporates diverse real-world scenes and topic-specific character interaction dialogues, thereby simulating characters' life experiences in various scenarios and topic-specific interactions with specific emotions. Experimental results indicate that our proposed framework achieves desirable performance and provides a valuable guideline for the construction of more accurate human simulacra in the future. Our data and code are publicly available at https://github.com/Bernard-Yang/SimsChat. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17962v3-abstract-full').style.display = 'none'; document.getElementById('2406.17962v3-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> 16 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 25 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/2406.17911">arXiv:2406.17911</a> <span> [<a href="https://arxiv.org/pdf/2406.17911">pdf</a>, <a href="https://arxiv.org/format/2406.17911">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"> X-ray Made Simple: Radiology Report Generation and Evaluation with Layman's Terms </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kun Zhao</a>, <a href="/search/cs?searchtype=author&query=Xiao%2C+C">Chenghao Xiao</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+C">Chen Tang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bohao Yang</a>, <a href="/search/cs?searchtype=author&query=Ye%2C+K">Kai Ye</a>, <a href="/search/cs?searchtype=author&query=Moubayed%2C+N+A">Noura Al Moubayed</a>, <a href="/search/cs?searchtype=author&query=Zhan%2C+L">Liang Zhan</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+C">Chenghua Lin</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="2406.17911v3-abstract-short" style="display: inline;"> Radiology Report Generation (RRG) has achieved significant progress with the advancements of multimodal generative models. However, the evaluation in the domain suffers from a lack of fair and robust metrics. We reveal that, high performance on RRG with existing lexical-based metrics (e.g. BLEU) might be more of a mirage - a model can get a high BLEU only by learning the template of reports. This… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17911v3-abstract-full').style.display = 'inline'; document.getElementById('2406.17911v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.17911v3-abstract-full" style="display: none;"> Radiology Report Generation (RRG) has achieved significant progress with the advancements of multimodal generative models. However, the evaluation in the domain suffers from a lack of fair and robust metrics. We reveal that, high performance on RRG with existing lexical-based metrics (e.g. BLEU) might be more of a mirage - a model can get a high BLEU only by learning the template of reports. This has become an urgent problem for RRG due to the highly patternized nature of these reports. In this work, we un-intuitively approach this problem by proposing the Layman's RRG framework, a layman's terms-based dataset, evaluation and training framework that systematically improves RRG with day-to-day language. We first contribute the translated Layman's terms dataset. Building upon the dataset, we then propose a semantics-based evaluation method, which is proved to mitigate the inflated numbers of BLEU and provides fairer evaluation. Last, we show that training on the layman's terms dataset encourages models to focus on the semantics of the reports, as opposed to overfitting to learning the report templates. We reveal a promising scaling law between the number of training examples and semantics gain provided by our dataset, compared to the inverse pattern brought by the original formats. Our code is available at \url{https://github.com/hegehongcha/LaymanRRG}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17911v3-abstract-full').style.display = 'none'; document.getElementById('2406.17911v3-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> 16 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 25 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/2406.17873">arXiv:2406.17873</a> <span> [<a href="https://arxiv.org/pdf/2406.17873">pdf</a>, <a href="https://arxiv.org/format/2406.17873">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"> Improving Arithmetic Reasoning Ability of Large Language Models through Relation Tuples, Verification and Dynamic Feedback </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Miao%2C+Z">Zhongtao Miao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiyan Zhao</a>, <a href="/search/cs?searchtype=author&query=Tsuruoka%2C+Y">Yoshimasa Tsuruoka</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="2406.17873v1-abstract-short" style="display: inline;"> Current representations used in reasoning steps of large language models can mostly be categorized into two main types: (1) natural language, which is difficult to verify; and (2) non-natural language, usually programming code, which is difficult for people who are unfamiliar with coding to read. In this paper, we propose to use a semi-structured form to represent reasoning steps of large language… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17873v1-abstract-full').style.display = 'inline'; document.getElementById('2406.17873v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.17873v1-abstract-full" style="display: none;"> Current representations used in reasoning steps of large language models can mostly be categorized into two main types: (1) natural language, which is difficult to verify; and (2) non-natural language, usually programming code, which is difficult for people who are unfamiliar with coding to read. In this paper, we propose to use a semi-structured form to represent reasoning steps of large language models. Specifically, we use relation tuples, which are not only human-readable but also machine-friendly and easier to verify than natural language. We implement a framework that includes three main components: (1) introducing relation tuples into the reasoning steps of large language models; (2) implementing an automatic verification process of reasoning steps with a local code interpreter based on relation tuples; and (3) integrating a simple and effective dynamic feedback mechanism, which we found helpful for self-improvement of large language models. The experimental results on various arithmetic datasets demonstrate the effectiveness of our method in improving the arithmetic reasoning ability of large language models. The source code is available at https://github.com/gpgg/art. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.17873v1-abstract-full').style.display = 'none'; document.getElementById('2406.17873v1-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> 25 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 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">Under review, 25 figures, 8 tables, 29 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.15758">arXiv:2406.15758</a> <span> [<a href="https://arxiv.org/pdf/2406.15758">pdf</a>, <a href="https://arxiv.org/format/2406.15758">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="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> EDGE-LLM: Enabling Efficient Large Language Model Adaptation on Edge Devices via Layerwise Unified Compression and Adaptive Layer Tuning and Voting </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yu%2C+Z">Zhongzhi Yu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zheng Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yuhan Li</a>, <a href="/search/cs?searchtype=author&query=You%2C+H">Haoran You</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+R">Ruijie Gao</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+X">Xiaoya Zhou</a>, <a href="/search/cs?searchtype=author&query=Bommu%2C+S+R">Sreenidhi Reedy Bommu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+Y+K">Yang Katie Zhao</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+Y+C">Yingyan Celine Lin</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="2406.15758v1-abstract-short" style="display: inline;"> Efficient adaption of large language models (LLMs) on edge devices is essential for applications requiring continuous and privacy-preserving adaptation and inference. However, existing tuning techniques fall short because of the high computation and memory overheads. To this end, we introduce a computation- and memory-efficient LLM tuning framework, called Edge-LLM, to facilitate affordable and ef… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15758v1-abstract-full').style.display = 'inline'; document.getElementById('2406.15758v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.15758v1-abstract-full" style="display: none;"> Efficient adaption of large language models (LLMs) on edge devices is essential for applications requiring continuous and privacy-preserving adaptation and inference. However, existing tuning techniques fall short because of the high computation and memory overheads. To this end, we introduce a computation- and memory-efficient LLM tuning framework, called Edge-LLM, to facilitate affordable and effective LLM adaptation on edge devices. Specifically, Edge-LLM features three core components: (1) a layer-wise unified compression (LUC) technique to reduce the computation overhead by generating layer-wise pruning sparsity and quantization bit-width policies, (2) an adaptive layer tuning and voting scheme to reduce the memory overhead by reducing the backpropagation depth, and (3) a complementary hardware scheduling strategy to handle the irregular computation patterns introduced by LUC and adaptive layer tuning, thereby achieving efficient computation and data movements. Extensive experiments demonstrate that Edge-LLM achieves a 2.92x speed up and a 4x memory overhead reduction as compared to vanilla tuning methods with comparable task accuracy. Our code is available at https://github.com/GATECH-EIC/Edge-LLM <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15758v1-abstract-full').style.display = 'none'; document.getElementById('2406.15758v1-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 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/2406.10311">arXiv:2406.10311</a> <span> [<a href="https://arxiv.org/pdf/2406.10311">pdf</a>, <a href="https://arxiv.org/format/2406.10311">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"> CHiSafetyBench: A Chinese Hierarchical Safety Benchmark for Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wenjing Zhang</a>, <a href="/search/cs?searchtype=author&query=Lei%2C+X">Xuejiao Lei</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Z">Zhaoxiang Liu</a>, <a href="/search/cs?searchtype=author&query=An%2C+M">Meijuan An</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bikun Yang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">KaiKai Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+K">Kai Wang</a>, <a href="/search/cs?searchtype=author&query=Lian%2C+S">Shiguo Lian</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="2406.10311v2-abstract-short" style="display: inline;"> With the profound development of large language models(LLMs), their safety concerns have garnered increasing attention. However, there is a scarcity of Chinese safety benchmarks for LLMs, and the existing safety taxonomies are inadequate, lacking comprehensive safety detection capabilities in authentic Chinese scenarios. In this work, we introduce CHiSafetyBench, a dedicated safety benchmark for e… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10311v2-abstract-full').style.display = 'inline'; document.getElementById('2406.10311v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.10311v2-abstract-full" style="display: none;"> With the profound development of large language models(LLMs), their safety concerns have garnered increasing attention. However, there is a scarcity of Chinese safety benchmarks for LLMs, and the existing safety taxonomies are inadequate, lacking comprehensive safety detection capabilities in authentic Chinese scenarios. In this work, we introduce CHiSafetyBench, a dedicated safety benchmark for evaluating LLMs' capabilities in identifying risky content and refusing answering risky questions in Chinese contexts. CHiSafetyBench incorporates a dataset that covers a hierarchical Chinese safety taxonomy consisting of 5 risk areas and 31 categories. This dataset comprises two types of tasks: multiple-choice questions and question-answering, evaluating LLMs from the perspectives of risk content identification and the ability to refuse answering risky questions respectively. Utilizing this benchmark, we validate the feasibility of automatic evaluation as a substitute for human evaluation and conduct comprehensive automatic safety assessments on mainstream Chinese LLMs. Our experiments reveal the varying performance of different models across various safety domains, indicating that all models possess considerable potential for improvement in Chinese safety capabilities. Our dataset is publicly available at https://github.com/UnicomAI/UnicomBenchmark/tree/main/CHiSafetyBench. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10311v2-abstract-full').style.display = 'none'; document.getElementById('2406.10311v2-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 14 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 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">16 pages, 5 figures</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.10307">arXiv:2406.10307</a> <span> [<a href="https://arxiv.org/pdf/2406.10307">pdf</a>, <a href="https://arxiv.org/format/2406.10307">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"> What is the best model? Application-driven Evaluation for Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lian%2C+S">Shiguo Lian</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaikai Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xinhui Liu</a>, <a href="/search/cs?searchtype=author&query=Lei%2C+X">Xuejiao Lei</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bikun Yang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wenjing Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+K">Kai Wang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Z">Zhaoxiang Liu</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="2406.10307v1-abstract-short" style="display: inline;"> General large language models enhanced with supervised fine-tuning and reinforcement learning from human feedback are increasingly popular in academia and industry as they generalize foundation models to various practical tasks in a prompt manner. To assist users in selecting the best model in practical application scenarios, i.e., choosing the model that meets the application requirements while m… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10307v1-abstract-full').style.display = 'inline'; document.getElementById('2406.10307v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.10307v1-abstract-full" style="display: none;"> General large language models enhanced with supervised fine-tuning and reinforcement learning from human feedback are increasingly popular in academia and industry as they generalize foundation models to various practical tasks in a prompt manner. To assist users in selecting the best model in practical application scenarios, i.e., choosing the model that meets the application requirements while minimizing cost, we introduce A-Eval, an application-driven LLMs evaluation benchmark for general large language models. First, we categorize evaluation tasks into five main categories and 27 sub-categories from a practical application perspective. Next, we construct a dataset comprising 678 question-and-answer pairs through a process of collecting, annotating, and reviewing. Then, we design an objective and effective evaluation method and evaluate a series of LLMs of different scales on A-Eval. Finally, we reveal interesting laws regarding model scale and task difficulty level and propose a feasible method for selecting the best model. Through A-Eval, we provide clear empirical and engineer guidance for selecting the best model, reducing barriers to selecting and using LLMs and promoting their application and development. Our benchmark is publicly available at https://github.com/UnicomAI/DataSet/tree/main/TestData/GeneralAbility. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.10307v1-abstract-full').style.display = 'none'; document.getElementById('2406.10307v1-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> 14 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/2406.09073">arXiv:2406.09073</a> <span> [<a href="https://arxiv.org/pdf/2406.09073">pdf</a>, <a href="https://arxiv.org/format/2406.09073">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> </div> </div> <p class="title is-5 mathjax"> Are we making progress in unlearning? Findings from the first NeurIPS unlearning competition </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Triantafillou%2C+E">Eleni Triantafillou</a>, <a href="/search/cs?searchtype=author&query=Kairouz%2C+P">Peter Kairouz</a>, <a href="/search/cs?searchtype=author&query=Pedregosa%2C+F">Fabian Pedregosa</a>, <a href="/search/cs?searchtype=author&query=Hayes%2C+J">Jamie Hayes</a>, <a href="/search/cs?searchtype=author&query=Kurmanji%2C+M">Meghdad Kurmanji</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kairan Zhao</a>, <a href="/search/cs?searchtype=author&query=Dumoulin%2C+V">Vincent Dumoulin</a>, <a href="/search/cs?searchtype=author&query=Junior%2C+J+J">Julio Jacques Junior</a>, <a href="/search/cs?searchtype=author&query=Mitliagkas%2C+I">Ioannis Mitliagkas</a>, <a href="/search/cs?searchtype=author&query=Wan%2C+J">Jun Wan</a>, <a href="/search/cs?searchtype=author&query=Hosoya%2C+L+S">Lisheng Sun Hosoya</a>, <a href="/search/cs?searchtype=author&query=Escalera%2C+S">Sergio Escalera</a>, <a href="/search/cs?searchtype=author&query=Dziugaite%2C+G+K">Gintare Karolina Dziugaite</a>, <a href="/search/cs?searchtype=author&query=Triantafillou%2C+P">Peter Triantafillou</a>, <a href="/search/cs?searchtype=author&query=Guyon%2C+I">Isabelle Guyon</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="2406.09073v1-abstract-short" style="display: inline;"> We present the findings of the first NeurIPS competition on unlearning, which sought to stimulate the development of novel algorithms and initiate discussions on formal and robust evaluation methodologies. The competition was highly successful: nearly 1,200 teams from across the world participated, and a wealth of novel, imaginative solutions with different characteristics were contributed. In thi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.09073v1-abstract-full').style.display = 'inline'; document.getElementById('2406.09073v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.09073v1-abstract-full" style="display: none;"> We present the findings of the first NeurIPS competition on unlearning, which sought to stimulate the development of novel algorithms and initiate discussions on formal and robust evaluation methodologies. The competition was highly successful: nearly 1,200 teams from across the world participated, and a wealth of novel, imaginative solutions with different characteristics were contributed. In this paper, we analyze top solutions and delve into discussions on benchmarking unlearning, which itself is a research problem. The evaluation methodology we developed for the competition measures forgetting quality according to a formal notion of unlearning, while incorporating model utility for a holistic evaluation. We analyze the effectiveness of different instantiations of this evaluation framework vis-a-vis the associated compute cost, and discuss implications for standardizing evaluation. We find that the ranking of leading methods remains stable under several variations of this framework, pointing to avenues for reducing the cost of evaluation. Overall, our findings indicate progress in unlearning, with top-performing competition entries surpassing existing algorithms under our evaluation framework. We analyze trade-offs made by different algorithms and strengths or weaknesses in terms of generalizability to new datasets, paving the way for advancing both benchmarking and algorithm development in this important area. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.09073v1-abstract-full').style.display = 'none'; document.getElementById('2406.09073v1-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> 13 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/2406.06600">arXiv:2406.06600</a> <span> [<a href="https://arxiv.org/pdf/2406.06600">pdf</a>, <a href="https://arxiv.org/format/2406.06600">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> HORAE: A Domain-Agnostic Modeling Language for Automating Multimodal Service Regulation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sun%2C+Y">Yutao Sun</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+M">Mingshuai Chen</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+T">Tiancheng Zhao</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kangjia Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+H">He Li</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+J">Jintao Chen</a>, <a href="/search/cs?searchtype=author&query=Lu%2C+L">Liqiang Lu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+X">Xinkui Zhao</a>, <a href="/search/cs?searchtype=author&query=Deng%2C+S">Shuiguang Deng</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+J">Jianwei Yin</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="2406.06600v2-abstract-short" style="display: inline;"> Artificial intelligence is rapidly encroaching on the field of service regulation. This work presents the design principles behind HORAE, a unified specification language to model multimodal regulation rules across a diverse set of domains. We show how HORAE facilitates an intelligent service regulation pipeline by further exploiting a fine-tuned large language model named HORAE that automates the… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.06600v2-abstract-full').style.display = 'inline'; document.getElementById('2406.06600v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.06600v2-abstract-full" style="display: none;"> Artificial intelligence is rapidly encroaching on the field of service regulation. This work presents the design principles behind HORAE, a unified specification language to model multimodal regulation rules across a diverse set of domains. We show how HORAE facilitates an intelligent service regulation pipeline by further exploiting a fine-tuned large language model named HORAE that automates the HORAE modeling process, thereby yielding an end-to-end framework for fully automated intelligent service regulation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.06600v2-abstract-full').style.display = 'none'; document.getElementById('2406.06600v2-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 6 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/2406.06571">arXiv:2406.06571</a> <span> [<a href="https://arxiv.org/pdf/2406.06571">pdf</a>, <a href="https://arxiv.org/format/2406.06571">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"> SUBLLM: A Novel Efficient Architecture with Token Sequence Subsampling for LLM </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Q">Quandong Wang</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+Y">Yuxuan Yuan</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+X">Xiaoyu Yang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+R">Ruike Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kang Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+W">Wei Liu</a>, <a href="/search/cs?searchtype=author&query=Luan%2C+J">Jian Luan</a>, <a href="/search/cs?searchtype=author&query=Povey%2C+D">Daniel Povey</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+B">Bin Wang</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="2406.06571v5-abstract-short" style="display: inline;"> While Large Language Models (LLMs) have achieved remarkable success in various fields, the efficiency of training and inference remains a major challenge. To address this issue, we propose SUBLLM, short for Subsampling-Upsampling-Bypass Large Language Model, an innovative architecture that extends the core decoder-only framework by incorporating subsampling, upsampling, and bypass modules. The sub… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.06571v5-abstract-full').style.display = 'inline'; document.getElementById('2406.06571v5-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.06571v5-abstract-full" style="display: none;"> While Large Language Models (LLMs) have achieved remarkable success in various fields, the efficiency of training and inference remains a major challenge. To address this issue, we propose SUBLLM, short for Subsampling-Upsampling-Bypass Large Language Model, an innovative architecture that extends the core decoder-only framework by incorporating subsampling, upsampling, and bypass modules. The subsampling modules are responsible for shortening the sequence, while the upsampling modules restore the sequence length, and the bypass modules enhance convergence. In comparison to LLaMA, the proposed SUBLLM exhibits significant enhancements in both training and inference speeds as well as memory usage, while maintaining competitive few-shot performance. During training, SUBLLM increases speeds by 26% and cuts memory by 10GB per GPU. In inference, it boosts speeds by up to 37% and reduces memory by 1GB per GPU. The training and inference speeds can be enhanced by 34% and 52% respectively when the context window is expanded to 8192. Our code is available at https://github.com/XiaoMi/subllm. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.06571v5-abstract-full').style.display = 'none'; document.getElementById('2406.06571v5-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 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 3 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 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">10 pages, 5 figures, accepted by ECAI 2024</span> </p> <p class="comments is-size-7"> <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/2406.03725">arXiv:2406.03725</a> <span> [<a href="https://arxiv.org/pdf/2406.03725">pdf</a>, <a href="https://arxiv.org/format/2406.03725">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"> LLMEmbed: Rethinking Lightweight LLM's Genuine Function in Text Classification </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Liu%2C+C">Chun Liu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Hongguang Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kainan Zhao</a>, <a href="/search/cs?searchtype=author&query=Ju%2C+X">Xinghai Ju</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+L">Lin Yang</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="2406.03725v1-abstract-short" style="display: inline;"> With the booming of Large Language Models (LLMs), prompt-learning has become a promising method mainly researched in various research areas. Recently, many attempts based on prompt-learning have been made to improve the performance of text classification. However, most of these methods are based on heuristic Chain-of-Thought (CoT), and tend to be more complex but less efficient. In this paper, we… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.03725v1-abstract-full').style.display = 'inline'; document.getElementById('2406.03725v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.03725v1-abstract-full" style="display: none;"> With the booming of Large Language Models (LLMs), prompt-learning has become a promising method mainly researched in various research areas. Recently, many attempts based on prompt-learning have been made to improve the performance of text classification. However, most of these methods are based on heuristic Chain-of-Thought (CoT), and tend to be more complex but less efficient. In this paper, we rethink the LLM-based text classification methodology, propose a simple and effective transfer learning strategy, namely LLMEmbed, to address this classical but challenging task. To illustrate, we first study how to properly extract and fuse the text embeddings via various lightweight LLMs at different network depths to improve their robustness and discrimination, then adapt such embeddings to train the classifier. We perform extensive experiments on publicly available datasets, and the results show that LLMEmbed achieves strong performance while enjoys low training overhead using lightweight LLM backbones compared to recent methods based on larger LLMs, i.e. GPT-3, and sophisticated prompt-based strategies. Our LLMEmbed achieves adequate accuracy on publicly available benchmarks without any fine-tuning while merely use 4% model parameters, 1.8% electricity consumption and 1.5% runtime compared to its counterparts. Code is available at: https://github.com/ChunLiu-cs/LLMEmbed-ACL2024. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.03725v1-abstract-full').style.display = 'none'; document.getElementById('2406.03725v1-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 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 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">ACL 2024 main conference</span> </p> </li> </ol> <nav class="pagination is-small is-centered breathe-horizontal" role="navigation" aria-label="pagination"> <a href="" class="pagination-previous is-invisible">Previous </a> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Zhao%2C+K&start=200" class="pagination-link " aria-label="Page 5" aria-current="page">5 </a> </li> <li><span class="pagination-ellipsis">…</span></li> </ul> </nav> <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