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 1,177 results for author: <span class="mathjax">Park, S</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=Park%2C+S">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="Park, S"> </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=Park%2C+S&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="Park, S"> <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=Park%2C+S&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&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/2503.19984">arXiv:2503.19984</a> <span> [<a href="https://arxiv.org/pdf/2503.19984">pdf</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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Fluid Dynamics">physics.flu-dyn</span> </div> </div> <p class="title is-5 mathjax"> Hybrid Magnetically and Electrically Powered Metallo-Dielectric Janus Microrobots: Enhanced Motion Control and Operation Beyond Planar Limits </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Rachbuch%2C+I">Ido Rachbuch</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sinwook Park</a>, <a href="/search/cs?searchtype=author&query=Katz%2C+Y">Yuval Katz</a>, <a href="/search/cs?searchtype=author&query=Miloh%2C+T">Touvia Miloh</a>, <a href="/search/cs?searchtype=author&query=Yossifon%2C+G">Gilad Yossifon</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="2503.19984v1-abstract-short" style="display: inline;"> This study introduces the integration of hybrid magnetic and electric actuation mechanisms to achieve advanced motion capabilities for Janus particle (JP) microrobots. We demonstrate enhanced in-plane motion control through versatile control strategies and present the concepts of interplanar transitions and 2.5-dimensional (2.5D) trajectories, enabled by magnetic levitation and electrostatic trapp… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.19984v1-abstract-full').style.display = 'inline'; document.getElementById('2503.19984v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.19984v1-abstract-full" style="display: none;"> This study introduces the integration of hybrid magnetic and electric actuation mechanisms to achieve advanced motion capabilities for Janus particle (JP) microrobots. We demonstrate enhanced in-plane motion control through versatile control strategies and present the concepts of interplanar transitions and 2.5-dimensional (2.5D) trajectories, enabled by magnetic levitation and electrostatic trapping. These innovations expand the mobility of JPs into 3D space, allowing dynamic operation beyond the limitations of traditional surface-bound motion. Key functionalities include obstacle crossing, transitions to elevated surfaces, and discrete surface patterning enabling highly localized interventions. Using this set of tools, we also showcase the controlled out-of-plane transport of both synthetic and biological cargo. Together, these advancements lay the groundwork for novel microrobot-related applications in microfluidic systems and biomedical research. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.19984v1-abstract-full').style.display = 'none'; document.getElementById('2503.19984v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.18339">arXiv:2503.18339</a> <span> [<a href="https://arxiv.org/pdf/2503.18339">pdf</a>, <a href="https://arxiv.org/format/2503.18339">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"> GranQ: Granular Zero-Shot Quantization with Unified Layer-Channel Awareness </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hong%2C+I">Inpyo Hong</a>, <a href="/search/cs?searchtype=author&query=Jo%2C+Y">Youngwan Jo</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+H">Hyojeong Lee</a>, <a href="/search/cs?searchtype=author&query=Ahn%2C+S">Sunghyun Ahn</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sanghyun Park</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="2503.18339v1-abstract-short" style="display: inline;"> Zero-shot quantization (ZSQ) enables neural network compression without training data, which is crucial in restricted data access environments. However, existing ZSQ methods suffer from significant activation loss in low-bit environments owing to their coarse-grained scaling strategy. To address this issue, we propose GranQ, a novel ZSQ approach that leverages layer-channel awareness to minimize t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.18339v1-abstract-full').style.display = 'inline'; document.getElementById('2503.18339v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.18339v1-abstract-full" style="display: none;"> Zero-shot quantization (ZSQ) enables neural network compression without training data, which is crucial in restricted data access environments. However, existing ZSQ methods suffer from significant activation loss in low-bit environments owing to their coarse-grained scaling strategy. To address this issue, we propose GranQ, a novel ZSQ approach that leverages layer-channel awareness to minimize the quantization error. Unlike conventional layer- or channel-wise quantization, GranQ dynamically adjusts quantization granularity by considering both layer- and channel-level activation distributions. This enables fine-grained quantization while minimizing activation distortion. Additionally, we introduce vectorized activation quantization, which enables efficient parallel computation and reduces computational overhead while preserving accuracy. GranQ achieves superior performance compared with those of state-of-the-art ZSQ methods that employ quantization-aware training. With these findings, we anticipate that GranQ will inspire novel research directions beyond conventional ZSQ approaches focused on data generation and model training. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.18339v1-abstract-full').style.display = 'none'; document.getElementById('2503.18339v1-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> 24 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.14831">arXiv:2503.14831</a> <span> [<a href="https://arxiv.org/pdf/2503.14831">pdf</a>, <a href="https://arxiv.org/format/2503.14831">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Signal Processing">eess.SP</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"> Robust Transmission of Punctured Text with Large Language Model-based Recovery </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sojeong Park</a>, <a href="/search/cs?searchtype=author&query=Noh%2C+H">Hyeonho Noh</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+H+J">Hyun Jong 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="2503.14831v1-abstract-short" style="display: inline;"> With the recent advancements in deep learning, semantic communication which transmits only task-oriented features, has rapidly emerged. However, since feature extraction relies on learning-based models, its performance fundamentally depends on the training dataset or tasks. For practical scenarios, it is essential to design a model that demonstrates robust performance regardless of dataset or task… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.14831v1-abstract-full').style.display = 'inline'; document.getElementById('2503.14831v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.14831v1-abstract-full" style="display: none;"> With the recent advancements in deep learning, semantic communication which transmits only task-oriented features, has rapidly emerged. However, since feature extraction relies on learning-based models, its performance fundamentally depends on the training dataset or tasks. For practical scenarios, it is essential to design a model that demonstrates robust performance regardless of dataset or tasks. In this correspondence, we propose a novel text transmission model that selects and transmits only a few characters and recovers the missing characters at the receiver using a large language model (LLM). Additionally, we propose a novel importance character extractor (ICE), which selects transmitted characters to enhance LLM recovery performance. Simulations demonstrate that the proposed filter selection by ICE outperforms random filter selection, which selects transmitted characters randomly. Moreover, the proposed model exhibits robust performance across different datasets and tasks and outperforms traditional bit-based communication in low signal-to-noise ratio conditions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.14831v1-abstract-full').style.display = 'none'; document.getElementById('2503.14831v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">This work has been submitted to the IEEE for possible publication</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.14453">arXiv:2503.14453</a> <span> [<a href="https://arxiv.org/pdf/2503.14453">pdf</a>, <a href="https://arxiv.org/format/2503.14453">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">stat.ML</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"> Online Conformal Probabilistic Numerics via Adaptive Edge-Cloud Offloading </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hou%2C+Q">Qiushuo Hou</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sangwoo Park</a>, <a href="/search/cs?searchtype=author&query=Zecchin%2C+M">Matteo Zecchin</a>, <a href="/search/cs?searchtype=author&query=Cai%2C+Y">Yunlong Cai</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+G">Guanding Yu</a>, <a href="/search/cs?searchtype=author&query=Simeone%2C+O">Osvaldo Simeone</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="2503.14453v1-abstract-short" style="display: inline;"> Consider an edge computing setting in which a user submits queries for the solution of a linear system to an edge processor, which is subject to time-varying computing availability. The edge processor applies a probabilistic linear solver (PLS) so as to be able to respond to the user's query within the allotted time and computing budget. Feedback to the user is in the form of an uncertainty set. D… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.14453v1-abstract-full').style.display = 'inline'; document.getElementById('2503.14453v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.14453v1-abstract-full" style="display: none;"> Consider an edge computing setting in which a user submits queries for the solution of a linear system to an edge processor, which is subject to time-varying computing availability. The edge processor applies a probabilistic linear solver (PLS) so as to be able to respond to the user's query within the allotted time and computing budget. Feedback to the user is in the form of an uncertainty set. Due to model misspecification, the uncertainty set obtained via a direct application of PLS does not come with coverage guarantees with respect to the true solution of the linear system. This work introduces a new method to calibrate the uncertainty sets produced by PLS with the aim of guaranteeing long-term coverage requirements. The proposed method, referred to as online conformal prediction-PLS (OCP-PLS), assumes sporadic feedback from cloud to edge. This enables the online calibration of uncertainty thresholds via online conformal prediction (OCP), an online optimization method previously studied in the context of prediction models. The validity of OCP-PLS is verified via experiments that bring insights into trade-offs between coverage, prediction set size, and cloud usage. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.14453v1-abstract-full').style.display = 'none'; document.getElementById('2503.14453v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">This paper has been submitted to a conference</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.12772">arXiv:2503.12772</a> <span> [<a href="https://arxiv.org/pdf/2503.12772">pdf</a>, <a href="https://arxiv.org/format/2503.12772">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> <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"> NuPlanQA: A Large-Scale Dataset and Benchmark for Multi-View Driving Scene Understanding in Multi-Modal Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sung-Yeon Park</a>, <a href="/search/cs?searchtype=author&query=Cui%2C+C">Can Cui</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+Y">Yunsheng Ma</a>, <a href="/search/cs?searchtype=author&query=Moradipari%2C+A">Ahmadreza Moradipari</a>, <a href="/search/cs?searchtype=author&query=Gupta%2C+R">Rohit Gupta</a>, <a href="/search/cs?searchtype=author&query=Han%2C+K">Kyungtae Han</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Ziran 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="2503.12772v1-abstract-short" style="display: inline;"> Recent advances in multi-modal large language models (MLLMs) have demonstrated strong performance across various domains; however, their ability to comprehend driving scenes remains less proven. The complexity of driving scenarios, which includes multi-view information, poses significant challenges for existing MLLMs. In this paper, we introduce NuPlanQA-Eval, a multi-view, multi-modal evaluation… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12772v1-abstract-full').style.display = 'inline'; document.getElementById('2503.12772v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.12772v1-abstract-full" style="display: none;"> Recent advances in multi-modal large language models (MLLMs) have demonstrated strong performance across various domains; however, their ability to comprehend driving scenes remains less proven. The complexity of driving scenarios, which includes multi-view information, poses significant challenges for existing MLLMs. In this paper, we introduce NuPlanQA-Eval, a multi-view, multi-modal evaluation benchmark for driving scene understanding. To further support generalization to multi-view driving scenarios, we also propose NuPlanQA-1M, a large-scale dataset comprising 1M real-world visual question-answering (VQA) pairs. For context-aware analysis of traffic scenes, we categorize our dataset into nine subtasks across three core skills: Road Environment Perception, Spatial Relations Recognition, and Ego-Centric Reasoning. Furthermore, we present BEV-LLM, integrating Bird's-Eye-View (BEV) features from multi-view images into MLLMs. Our evaluation results reveal key challenges that existing MLLMs face in driving scene-specific perception and spatial reasoning from ego-centric perspectives. In contrast, BEV-LLM demonstrates remarkable adaptability to this domain, outperforming other models in six of the nine subtasks. These findings highlight how BEV integration enhances multi-view MLLMs while also identifying key areas that require further refinement for effective adaptation to driving scenes. To facilitate further research, we publicly release NuPlanQA at https://github.com/sungyeonparkk/NuPlanQA. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12772v1-abstract-full').style.display = 'none'; document.getElementById('2503.12772v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.12524">arXiv:2503.12524</a> <span> [<a href="https://arxiv.org/pdf/2503.12524">pdf</a>, <a href="https://arxiv.org/format/2503.12524">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"> EXAONE Deep: Reasoning Enhanced Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Research%2C+L+A">LG AI Research</a>, <a href="/search/cs?searchtype=author&query=Bae%2C+K">Kyunghoon Bae</a>, <a href="/search/cs?searchtype=author&query=Choi%2C+E">Eunbi Choi</a>, <a href="/search/cs?searchtype=author&query=Choi%2C+K">Kibong Choi</a>, <a href="/search/cs?searchtype=author&query=Choi%2C+S+J">Stanley Jungkyu Choi</a>, <a href="/search/cs?searchtype=author&query=Choi%2C+Y">Yemuk Choi</a>, <a href="/search/cs?searchtype=author&query=Hong%2C+S">Seokhee Hong</a>, <a href="/search/cs?searchtype=author&query=Hwang%2C+J">Junwon Hwang</a>, <a href="/search/cs?searchtype=author&query=Jeon%2C+H">Hyojin Jeon</a>, <a href="/search/cs?searchtype=author&query=Jeon%2C+K">Kijeong Jeon</a>, <a href="/search/cs?searchtype=author&query=Jo%2C+G+J">Gerrard Jeongwon Jo</a>, <a href="/search/cs?searchtype=author&query=Jo%2C+H">Hyunjik Jo</a>, <a href="/search/cs?searchtype=author&query=Jung%2C+J">Jiyeon Jung</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+H">Hyosang Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+J">Joonkee Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Seonghwan Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Soyeon Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sunkyoung Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+Y">Yireun Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+Y">Yongil Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+Y">Youchul Kim</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+E+H">Edward Hwayoung Lee</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+H">Haeju Lee</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+H">Honglak Lee</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jinsik Lee</a> , et al. (7 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="2503.12524v2-abstract-short" style="display: inline;"> We present EXAONE Deep series, which exhibits superior capabilities in various reasoning tasks, including math and coding benchmarks. We train our models mainly on the reasoning-specialized dataset that incorporates long streams of thought processes. Evaluation results show that our smaller models, EXAONE Deep 2.4B and 7.8B, outperform other models of comparable size, while the largest model, EXAO… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12524v2-abstract-full').style.display = 'inline'; document.getElementById('2503.12524v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.12524v2-abstract-full" style="display: none;"> We present EXAONE Deep series, which exhibits superior capabilities in various reasoning tasks, including math and coding benchmarks. We train our models mainly on the reasoning-specialized dataset that incorporates long streams of thought processes. Evaluation results show that our smaller models, EXAONE Deep 2.4B and 7.8B, outperform other models of comparable size, while the largest model, EXAONE Deep 32B, demonstrates competitive performance against leading open-weight models. All EXAONE Deep models are openly available for research purposes and can be downloaded from https://huggingface.co/LGAI-EXAONE <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12524v2-abstract-full').style.display = 'none'; document.getElementById('2503.12524v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 16 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">arXiv admin note: substantial text overlap with arXiv:2412.04862, arXiv:2408.03541</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.12454">arXiv:2503.12454</a> <span> [<a href="https://arxiv.org/pdf/2503.12454">pdf</a>, <a href="https://arxiv.org/format/2503.12454">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="Computation">stat.CO</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">stat.ML</span> </div> </div> <p class="title is-5 mathjax"> Convergence Analysis of alpha-SVRG under Strong Convexity </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Xiao%2C+S">Sean Xiao</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sangwoo Park</a>, <a href="/search/cs?searchtype=author&query=Vlaski%2C+S">Stefan Vlaski</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="2503.12454v1-abstract-short" style="display: inline;"> Stochastic first-order methods for empirical risk minimization employ gradient approximations based on sampled data in lieu of exact gradients. Such constructions introduce noise into the learning dynamics, which can be corrected through variance-reduction techniques. There is increasing evidence in the literature that in many modern learning applications noise can have a beneficial effect on opti… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12454v1-abstract-full').style.display = 'inline'; document.getElementById('2503.12454v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.12454v1-abstract-full" style="display: none;"> Stochastic first-order methods for empirical risk minimization employ gradient approximations based on sampled data in lieu of exact gradients. Such constructions introduce noise into the learning dynamics, which can be corrected through variance-reduction techniques. There is increasing evidence in the literature that in many modern learning applications noise can have a beneficial effect on optimization and generalization. To this end, the recently proposed variance-reduction technique, alpha-SVRG [Yin et al., 2023] allows for fine-grained control of the level of residual noise in the learning dynamics, and has been reported to empirically outperform both SGD and SVRG in modern deep learning scenarios. By focusing on strongly convex environments, we first provide a unified convergence rate expression for alpha-SVRG under fixed learning rate, which reduces to that of either SGD or SVRG by setting alpha=0 or alpha=1, respectively. We show that alpha-SVRG has faster convergence rate compared to SGD and SVRG under suitable choice of alpha. Simulation results on linear regression validate our theory. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12454v1-abstract-full').style.display = 'none'; document.getElementById('2503.12454v1-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> 16 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">ICASSP 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.12297">arXiv:2503.12297</a> <span> [<a href="https://arxiv.org/pdf/2503.12297">pdf</a>, <a href="https://arxiv.org/format/2503.12297">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"> Train Robots in a JIF: Joint Inverse and Forward Dynamics with Human and Robot Demonstrations </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Khandate%2C+G">Gagan Khandate</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+B">Boxuan Wang</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sarah Park</a>, <a href="/search/cs?searchtype=author&query=Ni%2C+W">Weizhe Ni</a>, <a href="/search/cs?searchtype=author&query=Palacious%2C+J">Jaoquin Palacious</a>, <a href="/search/cs?searchtype=author&query=Lampo%2C+K">Kate Lampo</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+P">Philippe Wu</a>, <a href="/search/cs?searchtype=author&query=Ho%2C+R">Rosh Ho</a>, <a href="/search/cs?searchtype=author&query=Chang%2C+E">Eric Chang</a>, <a href="/search/cs?searchtype=author&query=Ciocarlie%2C+M">Matei Ciocarlie</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="2503.12297v1-abstract-short" style="display: inline;"> Pre-training on large datasets of robot demonstrations is a powerful technique for learning diverse manipulation skills but is often limited by the high cost and complexity of collecting robot-centric data, especially for tasks requiring tactile feedback. This work addresses these challenges by introducing a novel method for pre-training with multi-modal human demonstrations. Our approach jointly… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12297v1-abstract-full').style.display = 'inline'; document.getElementById('2503.12297v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.12297v1-abstract-full" style="display: none;"> Pre-training on large datasets of robot demonstrations is a powerful technique for learning diverse manipulation skills but is often limited by the high cost and complexity of collecting robot-centric data, especially for tasks requiring tactile feedback. This work addresses these challenges by introducing a novel method for pre-training with multi-modal human demonstrations. Our approach jointly learns inverse and forward dynamics to extract latent state representations, towards learning manipulation specific representations. This enables efficient fine-tuning with only a small number of robot demonstrations, significantly improving data efficiency. Furthermore, our method allows for the use of multi-modal data, such as combination of vision and touch for manipulation. By leveraging latent dynamics modeling and tactile sensing, this approach paves the way for scalable robot manipulation learning based on human demonstrations. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.12297v1-abstract-full').style.display = 'none'; document.getElementById('2503.12297v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 15 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">9 pages, 8 figures, submission to RSS 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.11315">arXiv:2503.11315</a> <span> [<a href="https://arxiv.org/pdf/2503.11315">pdf</a>, <a href="https://arxiv.org/format/2503.11315">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Sound">cs.SD</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"> MMS-LLaMA: Efficient LLM-based Audio-Visual Speech Recognition with Minimal Multimodal Speech Tokens </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yeo%2C+J+H">Jeong Hun Yeo</a>, <a href="/search/cs?searchtype=author&query=Rha%2C+H">Hyeongseop Rha</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S+J">Se Jin Park</a>, <a href="/search/cs?searchtype=author&query=Ro%2C+Y+M">Yong Man Ro</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="2503.11315v1-abstract-short" style="display: inline;"> Audio-Visual Speech Recognition (AVSR) achieves robust speech recognition in noisy environments by combining auditory and visual information. However, recent Large Language Model (LLM) based AVSR systems incur high computational costs due to the high temporal resolution of audio-visual speech processed by LLMs. In this work, we introduce an efficient multimodal speech LLM framework that minimizes… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.11315v1-abstract-full').style.display = 'inline'; document.getElementById('2503.11315v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.11315v1-abstract-full" style="display: none;"> Audio-Visual Speech Recognition (AVSR) achieves robust speech recognition in noisy environments by combining auditory and visual information. However, recent Large Language Model (LLM) based AVSR systems incur high computational costs due to the high temporal resolution of audio-visual speech processed by LLMs. In this work, we introduce an efficient multimodal speech LLM framework that minimizes token length while preserving essential linguistic content. Our approach employs an early av-fusion module for streamlined feature integration, an audio-visual speech Q-Former that dynamically allocates tokens based on input duration, and a refined query allocation strategy with a speech rate predictor to adjust token allocation according to speaking speed of each audio sample. Extensive experiments on the LRS3 dataset show that our method achieves state-of-the-art performance with a WER of 0.74% while using only 3.5 tokens per second. Moreover, our approach not only reduces token usage by 86% compared to the previous multimodal speech LLM framework, but also improves computational efficiency by reducing FLOPs by 35.7%. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.11315v1-abstract-full').style.display = 'none'; document.getElementById('2503.11315v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">The code and models are available https://github.com/JeongHun0716/MMS-LLaMA</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.10197">arXiv:2503.10197</a> <span> [<a href="https://arxiv.org/pdf/2503.10197">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Chemical Physics">physics.chem-ph</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 Chemical Reaction Outcomes Based on Electron Movements Using Machine Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+S">Shuan Chen</a>, <a href="/search/cs?searchtype=author&query=Park%2C+K+S">Kye Sung Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+T">Taewan Kim</a>, <a href="/search/cs?searchtype=author&query=Han%2C+S">Sunkyu Han</a>, <a href="/search/cs?searchtype=author&query=Jung%2C+Y">Yousung Jung</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="2503.10197v1-abstract-short" style="display: inline;"> Accurately predicting chemical reaction outcomes and potential byproducts is a fundamental task of modern chemistry, enabling the efficient design of synthetic pathways and driving progress in chemical science. Reaction mechanism, which tracks electron movements during chemical reactions, is critical for understanding reaction kinetics and identifying unexpected products. Here, we present Reactron… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.10197v1-abstract-full').style.display = 'inline'; document.getElementById('2503.10197v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.10197v1-abstract-full" style="display: none;"> Accurately predicting chemical reaction outcomes and potential byproducts is a fundamental task of modern chemistry, enabling the efficient design of synthetic pathways and driving progress in chemical science. Reaction mechanism, which tracks electron movements during chemical reactions, is critical for understanding reaction kinetics and identifying unexpected products. Here, we present Reactron, the first electron-based machine learning model for general reaction prediction. Reactron integrates electron movement into its predictions, generating detailed arrow-pushing diagrams that elucidate each mechanistic step leading to product formation. We demonstrate the high predictive performance of Reactron over existing product-only models by a large-scale reaction outcome prediction benchmark, and the adaptability of the model to learn new reactivity upon providing a few examples. Furthermore, it explores combinatorial reaction spaces, uncovering novel reactivities beyond its training data. With robust performance in both in- and out-of-distribution predictions, Reactron embodies human-like reasoning in chemistry and opens new frontiers in reaction discovery and synthesis design. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.10197v1-abstract-full').style.display = 'none'; document.getElementById('2503.10197v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">15 pages, 3 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/2503.09650">arXiv:2503.09650</a> <span> [<a href="https://arxiv.org/pdf/2503.09650">pdf</a>, <a href="https://arxiv.org/format/2503.09650">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Performance">cs.PF</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Hardware Architecture">cs.AR</span> </div> </div> <p class="title is-5 mathjax"> A Review on Proprietary Accelerators for Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sihyeong Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jemin Lee</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+B">Byung-Soo Kim</a>, <a href="/search/cs?searchtype=author&query=Jeon%2C+S">Seokhun Jeon</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="2503.09650v1-abstract-short" style="display: inline;"> With the advancement of Large Language Models (LLMs), the importance of accelerators that efficiently process LLM computations has been increasing. This paper discusses the necessity of LLM accelerators and provides a comprehensive analysis of the hardware and software characteristics of the main commercial LLM accelerators. Based on this analysis, we propose considerations for the development of… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.09650v1-abstract-full').style.display = 'inline'; document.getElementById('2503.09650v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.09650v1-abstract-full" style="display: none;"> With the advancement of Large Language Models (LLMs), the importance of accelerators that efficiently process LLM computations has been increasing. This paper discusses the necessity of LLM accelerators and provides a comprehensive analysis of the hardware and software characteristics of the main commercial LLM accelerators. Based on this analysis, we propose considerations for the development of next-generation LLM accelerators and suggest future research directions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.09650v1-abstract-full').style.display = 'none'; document.getElementById('2503.09650v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">4 pages, accepted in AICompS 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/2503.08092">arXiv:2503.08092</a> <span> [<a href="https://arxiv.org/pdf/2503.08092">pdf</a>, <a href="https://arxiv.org/format/2503.08092">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"> SparseVoxFormer: Sparse Voxel-based Transformer for Multi-modal 3D Object Detection </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Son%2C+H">Hyeongseok Son</a>, <a href="/search/cs?searchtype=author&query=He%2C+J">Jia He</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seung-In Park</a>, <a href="/search/cs?searchtype=author&query=Min%2C+Y">Ying Min</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yunhao Zhang</a>, <a href="/search/cs?searchtype=author&query=Yoo%2C+B">ByungIn Yoo</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="2503.08092v1-abstract-short" style="display: inline;"> Most previous 3D object detection methods that leverage the multi-modality of LiDAR and cameras utilize the Bird's Eye View (BEV) space for intermediate feature representation. However, this space uses a low x, y-resolution and sacrifices z-axis information to reduce the overall feature resolution, which may result in declined accuracy. To tackle the problem of using low-resolution features, this… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.08092v1-abstract-full').style.display = 'inline'; document.getElementById('2503.08092v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.08092v1-abstract-full" style="display: none;"> Most previous 3D object detection methods that leverage the multi-modality of LiDAR and cameras utilize the Bird's Eye View (BEV) space for intermediate feature representation. However, this space uses a low x, y-resolution and sacrifices z-axis information to reduce the overall feature resolution, which may result in declined accuracy. To tackle the problem of using low-resolution features, this paper focuses on the sparse nature of LiDAR point cloud data. From our observation, the number of occupied cells in the 3D voxels constructed from a LiDAR data can be even fewer than the number of total cells in the BEV map, despite the voxels' significantly higher resolution. Based on this, we introduce a novel sparse voxel-based transformer network for 3D object detection, dubbed as SparseVoxFormer. Instead of performing BEV feature extraction, we directly leverage sparse voxel features as the input for a transformer-based detector. Moreover, with regard to the camera modality, we introduce an explicit modality fusion approach that involves projecting 3D voxel coordinates onto 2D images and collecting the corresponding image features. Thanks to these components, our approach can leverage geometrically richer multi-modal features while even reducing the computational cost. Beyond the proof-of-concept level, we further focus on facilitating better multi-modal fusion and flexible control over the number of sparse features. Finally, thorough experimental results demonstrate that utilizing a significantly smaller number of sparse features drastically reduces computational costs in a 3D object detector while enhancing both overall and long-range performance. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.08092v1-abstract-full').style.display = 'none'; document.getElementById('2503.08092v1-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> 11 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.07216">arXiv:2503.07216</a> <span> [<a href="https://arxiv.org/pdf/2503.07216">pdf</a>, <a href="https://arxiv.org/format/2503.07216">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"> FedRand: Enhancing Privacy in Federated Learning with Randomized LoRA Subparameter Updates </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sangwoo Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+S">Seanie Lee</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+B">Byungjoo Kim</a>, <a href="/search/cs?searchtype=author&query=Hwang%2C+S+J">Sung Ju Hwang</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="2503.07216v2-abstract-short" style="display: inline;"> Federated Learning (FL) is a widely used framework for training models in a decentralized manner, ensuring that the central server does not have direct access to data from local clients. However, this approach may still fail to fully preserve data privacy, as models from local clients are exposed to the central server during the aggregation process. This issue becomes even more critical when train… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.07216v2-abstract-full').style.display = 'inline'; document.getElementById('2503.07216v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.07216v2-abstract-full" style="display: none;"> Federated Learning (FL) is a widely used framework for training models in a decentralized manner, ensuring that the central server does not have direct access to data from local clients. However, this approach may still fail to fully preserve data privacy, as models from local clients are exposed to the central server during the aggregation process. This issue becomes even more critical when training vision-language models (VLMs) with FL, as VLMs can easily memorize training data instances, making them vulnerable to membership inference attacks (MIAs). To address this challenge, we propose the FedRand framework, which avoids disclosing the full set of client parameters. In this framework, each client randomly selects subparameters of Low-Rank Adaptation (LoRA) from the server and keeps the remaining counterparts of the LoRA weights as private parameters. After training both parameters on the client's private dataset, only the non-private client parameters are sent back to the server for aggregation. This approach mitigates the risk of exposing client-side VLM parameters, thereby enhancing data privacy. We empirically validate that FedRand improves robustness against MIAs compared to relevant baselines while achieving accuracy comparable to methods that communicate full LoRA parameters across several benchmark datasets. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.07216v2-abstract-full').style.display = 'none'; document.getElementById('2503.07216v2-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> 11 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 10 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">Preprint</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.06300">arXiv:2503.06300</a> <span> [<a href="https://arxiv.org/pdf/2503.06300">pdf</a>, <a href="https://arxiv.org/format/2503.06300">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"> Efficient Gradient-Based Inference for Manipulation Planning in Contact Factor Graphs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jeongmin Lee</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sunkyung Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+M">Minji Lee</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+D">Dongjun 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="2503.06300v1-abstract-short" style="display: inline;"> This paper presents a framework designed to tackle a range of planning problems arise in manipulation, which typically involve complex geometric-physical reasoning related to contact and dynamic constraints. We introduce the Contact Factor Graph (CFG) to graphically model these diverse factors, enabling us to perform inference on the graphs to approximate the distribution and sample appropriate so… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.06300v1-abstract-full').style.display = 'inline'; document.getElementById('2503.06300v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.06300v1-abstract-full" style="display: none;"> This paper presents a framework designed to tackle a range of planning problems arise in manipulation, which typically involve complex geometric-physical reasoning related to contact and dynamic constraints. We introduce the Contact Factor Graph (CFG) to graphically model these diverse factors, enabling us to perform inference on the graphs to approximate the distribution and sample appropriate solutions. We propose a novel approach that can incorporate various phenomena of contact manipulation as differentiable factors, and develop an efficient inference algorithm for CFG that leverages this differentiability along with the conditional probabilities arising from the structured nature of contact. Our results demonstrate the capability of our framework in generating viable samples and approximating posterior distributions for various manipulation scenarios. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.06300v1-abstract-full').style.display = 'none'; document.getElementById('2503.06300v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 8 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">ICRA 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.05488">arXiv:2503.05488</a> <span> [<a href="https://arxiv.org/pdf/2503.05488">pdf</a>, <a href="https://arxiv.org/ps/2503.05488">ps</a>, <a href="https://arxiv.org/format/2503.05488">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"> KIEval: Evaluation Metric for Document Key Information Extraction </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Khang%2C+M">Minsoo Khang</a>, <a href="/search/cs?searchtype=author&query=Jung%2C+S+C">Sang Chul Jung</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sungrae Park</a>, <a href="/search/cs?searchtype=author&query=Hong%2C+T">Teakgyu Hong</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="2503.05488v2-abstract-short" style="display: inline;"> Document Key Information Extraction (KIE) is a technology that transforms valuable information in document images into structured data, and it has become an essential function in industrial settings. However, current evaluation metrics of this technology do not accurately reflect the critical attributes of its industrial applications. In this paper, we present KIEval, a novel application-centric e… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05488v2-abstract-full').style.display = 'inline'; document.getElementById('2503.05488v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.05488v2-abstract-full" style="display: none;"> Document Key Information Extraction (KIE) is a technology that transforms valuable information in document images into structured data, and it has become an essential function in industrial settings. However, current evaluation metrics of this technology do not accurately reflect the critical attributes of its industrial applications. In this paper, we present KIEval, a novel application-centric evaluation metric for Document KIE models. Unlike prior metrics, KIEval assesses Document KIE models not just on the extraction of individual information (entity) but also of the structured information (grouping). Evaluation of structured information provides assessment of Document KIE models that are more reflective of extracting grouped information from documents in industrial settings. Designed with industrial application in mind, we believe that KIEval can become a standard evaluation metric for developing or applying Document KIE models in practice. The code will be publicly available. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05488v2-abstract-full').style.display = 'none'; document.getElementById('2503.05488v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 7 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.05456">arXiv:2503.05456</a> <span> [<a href="https://arxiv.org/pdf/2503.05456">pdf</a>, <a href="https://arxiv.org/format/2503.05456">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</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/10.1145/3706598.3713530">10.1145/3706598.3713530 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> PinchCatcher: Enabling Multi-selection for Gaze+Pinch </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+J">Jinwook Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sangmin Park</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Q">Qiushi Zhou</a>, <a href="/search/cs?searchtype=author&query=Gonzalez-Franco%2C+M">Mar Gonzalez-Franco</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jeongmi Lee</a>, <a href="/search/cs?searchtype=author&query=Pfeuffer%2C+K">Ken Pfeuffer</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="2503.05456v2-abstract-short" style="display: inline;"> This paper investigates multi-selection in XR interfaces based on eye and hand interaction. We propose enabling multi-selection using different variations of techniques that combine gaze with a semi-pinch gesture, allowing users to select multiple objects, while on the way to a full-pinch. While our exploration is based on the semi-pinch mode for activating a quasi-mode, we explore four methods fo… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05456v2-abstract-full').style.display = 'inline'; document.getElementById('2503.05456v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.05456v2-abstract-full" style="display: none;"> This paper investigates multi-selection in XR interfaces based on eye and hand interaction. We propose enabling multi-selection using different variations of techniques that combine gaze with a semi-pinch gesture, allowing users to select multiple objects, while on the way to a full-pinch. While our exploration is based on the semi-pinch mode for activating a quasi-mode, we explore four methods for confirming subselections in multi-selection mode, varying in effort and complexity: dwell-time (SemiDwell), swipe (SemiSwipe), tilt (SemiTilt), and non-dominant hand input (SemiNDH), and compare them to a baseline technique. In the user study, we evaluate their effectiveness in reducing task completion time, errors, and effort. The results indicate the strengths and weaknesses of each technique, with SemiSwipe and SemiDwell as the most preferred methods by participants. We also demonstrate their utility in file managing and RTS gaming application scenarios. This study provides valuable insights to advance 3D input systems in XR. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05456v2-abstract-full').style.display = 'none'; document.getElementById('2503.05456v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 7 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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, CHI Conference on Human Factors in Computing Systems, 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.05186">arXiv:2503.05186</a> <span> [<a href="https://arxiv.org/pdf/2503.05186">pdf</a>, <a href="https://arxiv.org/format/2503.05186">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"> Narrating the Video: Boosting Text-Video Retrieval via Comprehensive Utilization of Frame-Level Captions </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hur%2C+C">Chan Hur</a>, <a href="/search/cs?searchtype=author&query=Hong%2C+J">Jeong-hun Hong</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+D">Dong-hun Lee</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+D">Dabin Kang</a>, <a href="/search/cs?searchtype=author&query=Myeong%2C+S">Semin Myeong</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sang-hyo Park</a>, <a href="/search/cs?searchtype=author&query=Park%2C+H">Hyeyoung Park</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="2503.05186v4-abstract-short" style="display: inline;"> In recent text-video retrieval, the use of additional captions from vision-language models has shown promising effects on the performance. However, existing models using additional captions often have struggled to capture the rich semantics, including temporal changes, inherent in the video. In addition, incorrect information caused by generative models can lead to inaccurate retrieval. To address… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05186v4-abstract-full').style.display = 'inline'; document.getElementById('2503.05186v4-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.05186v4-abstract-full" style="display: none;"> In recent text-video retrieval, the use of additional captions from vision-language models has shown promising effects on the performance. However, existing models using additional captions often have struggled to capture the rich semantics, including temporal changes, inherent in the video. In addition, incorrect information caused by generative models can lead to inaccurate retrieval. To address these issues, we propose a new framework, Narrating the Video (NarVid), which strategically leverages the comprehensive information available from frame-level captions, the narration. The proposed NarVid exploits narration in multiple ways: 1) feature enhancement through cross-modal interactions between narration and video, 2) query-aware adaptive filtering to suppress irrelevant or incorrect information, 3) dual-modal matching score by adding query-video similarity and query-narration similarity, and 4) hard-negative loss to learn discriminative features from multiple perspectives using the two similarities from different views. Experimental results demonstrate that NarVid achieves state-of-the-art performance on various benchmark datasets. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.05186v4-abstract-full').style.display = 'none'; document.getElementById('2503.05186v4-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 7 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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 at CVPR 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.04675">arXiv:2503.04675</a> <span> [<a href="https://arxiv.org/pdf/2503.04675">pdf</a>, <a href="https://arxiv.org/format/2503.04675">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"> LLM-guided Plan and Retrieval: A Strategic Alignment for Interpretable User Satisfaction Estimation in Dialogue </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sangyeop Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sohhyung Park</a>, <a href="/search/cs?searchtype=author&query=Jung%2C+J">Jaewon Jung</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+J">Jinseok Kim</a>, <a href="/search/cs?searchtype=author&query=Cho%2C+S">Sungzoon Cho</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="2503.04675v1-abstract-short" style="display: inline;"> Understanding user satisfaction with conversational systems, known as User Satisfaction Estimation (USE), is essential for assessing dialogue quality and enhancing user experiences. However, existing methods for USE face challenges due to limited understanding of underlying reasons for user dissatisfaction and the high costs of annotating user intentions. To address these challenges, we propose PR… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.04675v1-abstract-full').style.display = 'inline'; document.getElementById('2503.04675v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.04675v1-abstract-full" style="display: none;"> Understanding user satisfaction with conversational systems, known as User Satisfaction Estimation (USE), is essential for assessing dialogue quality and enhancing user experiences. However, existing methods for USE face challenges due to limited understanding of underlying reasons for user dissatisfaction and the high costs of annotating user intentions. To address these challenges, we propose PRAISE (Plan and Retrieval Alignment for Interpretable Satisfaction Estimation), an interpretable framework for effective user satisfaction prediction. PRAISE operates through three key modules. The Strategy Planner develops strategies, which are natural language criteria for classifying user satisfaction. The Feature Retriever then incorporates knowledge on user satisfaction from Large Language Models (LLMs) and retrieves relevance features from utterances. Finally, the Score Analyzer evaluates strategy predictions and classifies user satisfaction. Experimental results demonstrate that PRAISE achieves state-of-the-art performance on three benchmarks for the USE task. Beyond its superior performance, PRAISE offers additional benefits. It enhances interpretability by providing instance-level explanations through effective alignment of utterances with strategies. Moreover, PRAISE operates more efficiently than existing approaches by eliminating the need for LLMs during the inference phase. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.04675v1-abstract-full').style.display = 'none'; document.getElementById('2503.04675v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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 NAACL 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.04504">arXiv:2503.04504</a> <span> [<a href="https://arxiv.org/pdf/2503.04504">pdf</a>, <a href="https://arxiv.org/format/2503.04504">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"> AnyAnomaly: Zero-Shot Customizable Video Anomaly Detection with LVLM </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ahn%2C+S">Sunghyun Ahn</a>, <a href="/search/cs?searchtype=author&query=Jo%2C+Y">Youngwan Jo</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+K">Kijung Lee</a>, <a href="/search/cs?searchtype=author&query=Kwon%2C+S">Sein Kwon</a>, <a href="/search/cs?searchtype=author&query=Hong%2C+I">Inpyo Hong</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sanghyun Park</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="2503.04504v1-abstract-short" style="display: inline;"> Video anomaly detection (VAD) is crucial for video analysis and surveillance in computer vision. However, existing VAD models rely on learned normal patterns, which makes them difficult to apply to diverse environments. Consequently, users should retrain models or develop separate AI models for new environments, which requires expertise in machine learning, high-performance hardware, and extensive… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.04504v1-abstract-full').style.display = 'inline'; document.getElementById('2503.04504v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.04504v1-abstract-full" style="display: none;"> Video anomaly detection (VAD) is crucial for video analysis and surveillance in computer vision. However, existing VAD models rely on learned normal patterns, which makes them difficult to apply to diverse environments. Consequently, users should retrain models or develop separate AI models for new environments, which requires expertise in machine learning, high-performance hardware, and extensive data collection, limiting the practical usability of VAD. To address these challenges, this study proposes customizable video anomaly detection (C-VAD) technique and the AnyAnomaly model. C-VAD considers user-defined text as an abnormal event and detects frames containing a specified event in a video. We effectively implemented AnyAnomaly using a context-aware visual question answering without fine-tuning the large vision language model. To validate the effectiveness of the proposed model, we constructed C-VAD datasets and demonstrated the superiority of AnyAnomaly. Furthermore, our approach showed competitive performance on VAD benchmark datasets, achieving state-of-the-art results on the UBnormal dataset and outperforming other methods in generalization across all datasets. Our code is available online at github.com/SkiddieAhn/Paper-AnyAnomaly. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.04504v1-abstract-full').style.display = 'none'; document.getElementById('2503.04504v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.03974">arXiv:2503.03974</a> <span> [<a href="https://arxiv.org/pdf/2503.03974">pdf</a>, <a href="https://arxiv.org/format/2503.03974">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</span> </div> </div> <p class="title is-5 mathjax"> Cryptographic Verifiability for Voter Registration Systems </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=F%C3%A1brega%2C+A">Andr茅s F谩brega</a>, <a href="/search/cs?searchtype=author&query=Cable%2C+J">Jack Cable</a>, <a href="/search/cs?searchtype=author&query=Specter%2C+M+A">Michael A. Specter</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sunoo Park</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="2503.03974v1-abstract-short" style="display: inline;"> Voter registration systems are a critical - and surprisingly understudied - element of most high-stakes elections. Despite a history of targeting by adversaries, relatively little academic work has been done to increase visibility into how voter registration systems keep voters' data secure, accurate, and up to date. Enhancing transparency and verifiability could help election officials and the pu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.03974v1-abstract-full').style.display = 'inline'; document.getElementById('2503.03974v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.03974v1-abstract-full" style="display: none;"> Voter registration systems are a critical - and surprisingly understudied - element of most high-stakes elections. Despite a history of targeting by adversaries, relatively little academic work has been done to increase visibility into how voter registration systems keep voters' data secure, accurate, and up to date. Enhancing transparency and verifiability could help election officials and the public detect and mitigate risks to this essential component of electoral processes worldwide. This work introduces cryptographic verifiability for voter registration systems. Based on consultation with diverse expert stakeholders that support elections systems, we precisely define the requirements for cryptographic verifiability in voter registration and systematize the practical challenges that must be overcome for near-term deployment. We then introduce VRLog, the first system to bring strong verifiability to voter registration. VRLog enables election officials to provide a transparent log that (1) allows voters to verify that their registration data has not been tampered with and (2) allows the public to monitor update patterns and database consistency. We also introduce VRLog$^x$, an enhancement to VRLog that offers cryptographic privacy to voter deduplication between jurisdictions - a common maintenance task currently performed in plaintext or using trusted third parties. Our designs rely on standard, efficient cryptographic primitives, and are backward compatible with existing voter registration systems. Finally, we provide an open-source implementation of VRLog and benchmarks to demonstrate that the system is practical - capable of running on low-cost commodity hardware and scaling to support databases the size of the largest U.S. state voter registration systems. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.03974v1-abstract-full').style.display = 'none'; document.getElementById('2503.03974v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.03967">arXiv:2503.03967</a> <span> [<a href="https://arxiv.org/pdf/2503.03967">pdf</a>, <a href="https://arxiv.org/format/2503.03967">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</span> </div> </div> <p class="title is-5 mathjax"> Model Behavior Specification by Leveraging LLM Self-Playing and Self-Improving </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Soya Park</a>, <a href="/search/cs?searchtype=author&query=Zamfirescu-Pereira%2C+J+D">J. D. Zamfirescu-Pereira</a>, <a href="/search/cs?searchtype=author&query=Kulkarni%2C+C">Chinmay Kulkarni</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="2503.03967v1-abstract-short" style="display: inline;"> Training AI models is challenging, particularly when crafting behavior instructions. Traditional methods rely on machines (supervised learning) or manual pattern discovery, which results in not interpretable models or time sink. While Large Language Models (LLMs) simplify instruction writing through natural language, articulating intended model behavior still remains difficult. We introduce Visi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.03967v1-abstract-full').style.display = 'inline'; document.getElementById('2503.03967v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.03967v1-abstract-full" style="display: none;"> Training AI models is challenging, particularly when crafting behavior instructions. Traditional methods rely on machines (supervised learning) or manual pattern discovery, which results in not interpretable models or time sink. While Large Language Models (LLMs) simplify instruction writing through natural language, articulating intended model behavior still remains difficult. We introduce Visionary Tuning, a human-in-the-loop self-playing followed by automatic self-refinement to improve behavior specification. Our system helps users clarify desired behavior through self-playing and generates prompts through self-improving, Our first evaluation involves user study conducted on a system implementation of Visionary Tuning within the context of chatbot behavior. Our system self-play itself by simulating user interactions to identify patterns and create effective prompts based on the pattern. In a within-subject study (N=12), participants pinpointed more patterns through self-playing and crafted better prompts. Surprisingly, users felt more or less success level in specifying the model behavior. Follow-up crowd studies (N=60) confirmed that the chatbot adhered to instructions without sacrificing quality. Our second evaluation is a case study on a real-world implementation using a movie rating dataset with Visionary Tuning, demonstrating its effectiveness and robustness in modeling a critic's preferences across the spectrum of low to highly rated movies. Together, these results suggest how AI improves the design process of interactive AI systems. Furthermore, they suggest how the benefits of these tools may be non-obvious to end-users. We reflect on these findings and suggest future directions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.03967v1-abstract-full').style.display = 'none'; document.getElementById('2503.03967v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.01917">arXiv:2503.01917</a> <span> [<a href="https://arxiv.org/pdf/2503.01917">pdf</a>, <a href="https://arxiv.org/format/2503.01917">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"> How to Steer LLM Latents for Hallucination Detection? </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Seongheon Park</a>, <a href="/search/cs?searchtype=author&query=Du%2C+X">Xuefeng Du</a>, <a href="/search/cs?searchtype=author&query=Yeh%2C+M">Min-Hsuan Yeh</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Haobo Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yixuan 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="2503.01917v1-abstract-short" style="display: inline;"> Hallucinations in LLMs pose a significant concern to their safe deployment in real-world applications. Recent approaches have leveraged the latent space of LLMs for hallucination detection, but their embeddings, optimized for linguistic coherence rather than factual accuracy, often fail to clearly separate truthful and hallucinated content. To this end, we propose the Truthfulness Separator Vector… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01917v1-abstract-full').style.display = 'inline'; document.getElementById('2503.01917v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.01917v1-abstract-full" style="display: none;"> Hallucinations in LLMs pose a significant concern to their safe deployment in real-world applications. Recent approaches have leveraged the latent space of LLMs for hallucination detection, but their embeddings, optimized for linguistic coherence rather than factual accuracy, often fail to clearly separate truthful and hallucinated content. To this end, we propose the Truthfulness Separator Vector (TSV), a lightweight and flexible steering vector that reshapes the LLM's representation space during inference to enhance the separation between truthful and hallucinated outputs, without altering model parameters. Our two-stage framework first trains TSV on a small set of labeled exemplars to form compact and well-separated clusters. It then augments the exemplar set with unlabeled LLM generations, employing an optimal transport-based algorithm for pseudo-labeling combined with a confidence-based filtering process. Extensive experiments demonstrate that TSV achieves state-of-the-art performance with minimal labeled data, exhibiting strong generalization across datasets and providing a practical solution for real-world LLM applications. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01917v1-abstract-full').style.display = 'none'; document.getElementById('2503.01917v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">ICLR Workshop on Quantify Uncertainty and Hallucination in Foundation Models (QUESTION), 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.01216">arXiv:2503.01216</a> <span> [<a href="https://arxiv.org/pdf/2503.01216">pdf</a>, <a href="https://arxiv.org/format/2503.01216">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"> A Single Scale Doesn't Fit All: Adaptive Motion Scaling for Efficient and Precise Teleoperation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yoon%2C+J">Jeonghyeon Yoon</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sanghyeok Park</a>, <a href="/search/cs?searchtype=author&query=Park%2C+H">Hyojae Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+C">Cholin Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sihyeoung Park</a>, <a href="/search/cs?searchtype=author&query=Hwang%2C+M">Minho Hwang</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="2503.01216v1-abstract-short" style="display: inline;"> Teleoperation is increasingly employed in environments where direct human access is difficult, such as hazardous exploration or surgical field. However, if the motion scale factor(MSF) intended to compensate for workspace-size differences is set inappropriately, repeated clutching operations and reduced precision can significantly raise cognitive load. This paper presents a shared controller that… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01216v1-abstract-full').style.display = 'inline'; document.getElementById('2503.01216v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.01216v1-abstract-full" style="display: none;"> Teleoperation is increasingly employed in environments where direct human access is difficult, such as hazardous exploration or surgical field. However, if the motion scale factor(MSF) intended to compensate for workspace-size differences is set inappropriately, repeated clutching operations and reduced precision can significantly raise cognitive load. This paper presents a shared controller that dynamically applies the MSF based on the user's intended motion scale. Inspired by human motor skills, the leader arm trajectory is divided into coarse(fast, large-range movements) and fine(precise, small-range movements), with three features extracted to train a fuzzy C-means(FCM) clustering model that probabilistically classifies the user's motion scale. Scaling the robot's motion accordingly reduces unnecessary repetition for large-scale movements and enables more precise control for fine operations. Incorporating recent trajectory data into model updates and offering user feedback for adjusting the MSF range and response speed allows mutual adaptation between user and system. In peg transfer experiments, compared to using a fixed single scale, the proposed approach demonstrated improved task efficiency(number of clutching and task completion time decreased 38.46% and 11.96% respectively), while NASA-TLX scores confirmed a meaningful reduction(58.01% decreased) in cognitive load. This outcome suggests that a user-intent-based motion scale adjustment can effectively enhance both efficiency and precision in teleoperation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01216v1-abstract-full').style.display = 'none'; document.getElementById('2503.01216v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 3 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.01097">arXiv:2503.01097</a> <span> [<a href="https://arxiv.org/pdf/2503.01097">pdf</a>, <a href="https://arxiv.org/format/2503.01097">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"> Measuring the Validity of Clustering Validation Datasets </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jeon%2C+H">Hyeon Jeon</a>, <a href="/search/cs?searchtype=author&query=Aupetit%2C+M">Micha毛l Aupetit</a>, <a href="/search/cs?searchtype=author&query=Shin%2C+D">DongHwa Shin</a>, <a href="/search/cs?searchtype=author&query=Cho%2C+A">Aeri Cho</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seokhyeon Park</a>, <a href="/search/cs?searchtype=author&query=Seo%2C+J">Jinwook Seo</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="2503.01097v1-abstract-short" style="display: inline;"> Clustering techniques are often validated using benchmark datasets where class labels are used as ground-truth clusters. However, depending on the datasets, class labels may not align with the actual data clusters, and such misalignment hampers accurate validation. Therefore, it is essential to evaluate and compare datasets regarding their cluster-label matching (CLM), i.e., how well their class l… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01097v1-abstract-full').style.display = 'inline'; document.getElementById('2503.01097v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.01097v1-abstract-full" style="display: none;"> Clustering techniques are often validated using benchmark datasets where class labels are used as ground-truth clusters. However, depending on the datasets, class labels may not align with the actual data clusters, and such misalignment hampers accurate validation. Therefore, it is essential to evaluate and compare datasets regarding their cluster-label matching (CLM), i.e., how well their class labels match actual clusters. Internal validation measures (IVMs), like Silhouette, can compare CLM over different labeling of the same dataset, but are not designed to do so across different datasets. We thus introduce Adjusted IVMs as fast and reliable methods to evaluate and compare CLM across datasets. We establish four axioms that require validation measures to be independent of data properties not related to cluster structure (e.g., dimensionality, dataset size). Then, we develop standardized protocols to convert any IVM to satisfy these axioms, and use these protocols to adjust six widely used IVMs. Quantitative experiments (1) verify the necessity and effectiveness of our protocols and (2) show that adjusted IVMs outperform the competitors, including standard IVMs, in accurately evaluating CLM both within and across datasets. We also show that the datasets can be filtered or improved using our method to form more reliable benchmarks for clustering validation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.01097v1-abstract-full').style.display = 'none'; document.getElementById('2503.01097v1-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </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">IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2503.00032">arXiv:2503.00032</a> <span> [<a href="https://arxiv.org/pdf/2503.00032">pdf</a>, <a href="https://arxiv.org/format/2503.00032">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"> Detecting LLM-Generated Korean Text through Linguistic Feature Analysis </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Shinwoo Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Shubin Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+D">Do-Kyung Kim</a>, <a href="/search/cs?searchtype=author&query=Han%2C+Y">Yo-Sub Han</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="2503.00032v2-abstract-short" style="display: inline;"> The rapid advancement of large language models (LLMs) increases the difficulty of distinguishing between human-written and LLM-generated text. Detecting LLM-generated text is crucial for upholding academic integrity, preventing plagiarism, protecting copyrights, and ensuring ethical research practices. Most prior studies on detecting LLM-generated text focus primarily on English text. However, lan… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.00032v2-abstract-full').style.display = 'inline'; document.getElementById('2503.00032v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2503.00032v2-abstract-full" style="display: none;"> The rapid advancement of large language models (LLMs) increases the difficulty of distinguishing between human-written and LLM-generated text. Detecting LLM-generated text is crucial for upholding academic integrity, preventing plagiarism, protecting copyrights, and ensuring ethical research practices. Most prior studies on detecting LLM-generated text focus primarily on English text. However, languages with distinct morphological and syntactic characteristics require specialized detection approaches. Their unique structures and usage patterns can hinder the direct application of methods primarily designed for English. Among such languages, we focus on Korean, which has relatively flexible spacing rules, a rich morphological system, and less frequent comma usage compared to English. We introduce KatFish, the first benchmark dataset for detecting LLM-generated Korean text. The dataset consists of text written by humans and generated by four LLMs across three genres. By examining spacing patterns, part-of-speech diversity, and comma usage, we illuminate the linguistic differences between human-written and LLM-generated Korean text. Building on these observations, we propose KatFishNet, a detection method specifically designed for the Korean language. KatFishNet achieves an average of 19.78% higher AUROC compared to the best-performing existing detection method. Our code and data are available at https://github.com/Shinwoo-Park/detecting_llm_generated_korean_text_through_linguistic_analysis. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2503.00032v2-abstract-full').style.display = 'none'; document.getElementById('2503.00032v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 24 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> March 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.19930">arXiv:2502.19930</a> <span> [<a href="https://arxiv.org/pdf/2502.19930">pdf</a>, <a href="https://arxiv.org/format/2502.19930">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"> Identity-preserving Distillation Sampling by Fixed-Point Iterator </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+S">SeonHwa Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+J">Jiwon Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Soobin Park</a>, <a href="/search/cs?searchtype=author&query=Ahn%2C+D">Donghoon Ahn</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+J">Jiwon Kang</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Seungryong Kim</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+K+H">Kyong Hwan Jin</a>, <a href="/search/cs?searchtype=author&query=Cha%2C+E">Eunju Cha</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="2502.19930v2-abstract-short" style="display: inline;"> Score distillation sampling (SDS) demonstrates a powerful capability for text-conditioned 2D image and 3D object generation by distilling the knowledge from learned score functions. However, SDS often suffers from blurriness caused by noisy gradients. When SDS meets the image editing, such degradations can be reduced by adjusting bias shifts using reference pairs, but the de-biasing techniques are… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.19930v2-abstract-full').style.display = 'inline'; document.getElementById('2502.19930v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.19930v2-abstract-full" style="display: none;"> Score distillation sampling (SDS) demonstrates a powerful capability for text-conditioned 2D image and 3D object generation by distilling the knowledge from learned score functions. However, SDS often suffers from blurriness caused by noisy gradients. When SDS meets the image editing, such degradations can be reduced by adjusting bias shifts using reference pairs, but the de-biasing techniques are still corrupted by erroneous gradients. To this end, we introduce Identity-preserving Distillation Sampling (IDS), which compensates for the gradient leading to undesired changes in the results. Based on the analysis that these errors come from the text-conditioned scores, a new regularization technique, called fixed-point iterative regularization (FPR), is proposed to modify the score itself, driving the preservation of the identity even including poses and structures. Thanks to a self-correction by FPR, the proposed method provides clear and unambiguous representations corresponding to the given prompts in image-to-image editing and editable neural radiance field (NeRF). The structural consistency between the source and the edited data is obviously maintained compared to other state-of-the-art methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.19930v2-abstract-full').style.display = 'none'; document.getElementById('2502.19930v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 27 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.19759">arXiv:2502.19759</a> <span> [<a href="https://arxiv.org/pdf/2502.19759">pdf</a>, <a href="https://arxiv.org/format/2502.19759">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="Audio and Speech Processing">eess.AS</span> </div> </div> <p class="title is-5 mathjax"> Does Your Voice Assistant Remember? Analyzing Conversational Context Recall and Utilization in Voice Interaction Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+H">Heeseung Kim</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+C+H">Che Hyun Lee</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sangkwon Park</a>, <a href="/search/cs?searchtype=author&query=Yeom%2C+J">Jiheum Yeom</a>, <a href="/search/cs?searchtype=author&query=Park%2C+N">Nohil Park</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+S">Sangwon Yu</a>, <a href="/search/cs?searchtype=author&query=Yoon%2C+S">Sungroh Yoon</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="2502.19759v1-abstract-short" style="display: inline;"> Recent advancements in multi-turn voice interaction models have improved user-model communication. However, while closed-source models effectively retain and recall past utterances, whether open-source models share this ability remains unexplored. To fill this gap, we systematically evaluate how well open-source interaction models utilize past utterances using ContextDialog, a benchmark we propose… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.19759v1-abstract-full').style.display = 'inline'; document.getElementById('2502.19759v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.19759v1-abstract-full" style="display: none;"> Recent advancements in multi-turn voice interaction models have improved user-model communication. However, while closed-source models effectively retain and recall past utterances, whether open-source models share this ability remains unexplored. To fill this gap, we systematically evaluate how well open-source interaction models utilize past utterances using ContextDialog, a benchmark we proposed for this purpose. Our findings show that speech-based models have more difficulty than text-based ones, especially when recalling information conveyed in speech, and even with retrieval-augmented generation, models still struggle with questions about past utterances. These insights highlight key limitations in open-source models and suggest ways to improve memory retention and retrieval robustness. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.19759v1-abstract-full').style.display = 'none'; document.getElementById('2502.19759v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">Work in Progress, Project Page: https://contextdialog.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/2502.18853">arXiv:2502.18853</a> <span> [<a href="https://arxiv.org/pdf/2502.18853">pdf</a>, <a href="https://arxiv.org/format/2502.18853">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</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"> Reimagining Personal Data: Unlocking the Potential of AI-Generated Images in Personal Data Meaning-Making </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Soobin Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+H">Hankyung Kim</a>, <a href="/search/cs?searchtype=author&query=Lim%2C+Y">Youn-kyung Lim</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="2502.18853v1-abstract-short" style="display: inline;"> Image-generative AI provides new opportunities to transform personal data into alternative visual forms. In this paper, we illustrate the potential of AI-generated images in facilitating meaningful engagement with personal data. In a formative autobiographical design study, we explored the design and use of AI-generated images derived from personal data. Informed by this study, we designed a web-b… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18853v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18853v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18853v1-abstract-full" style="display: none;"> Image-generative AI provides new opportunities to transform personal data into alternative visual forms. In this paper, we illustrate the potential of AI-generated images in facilitating meaningful engagement with personal data. In a formative autobiographical design study, we explored the design and use of AI-generated images derived from personal data. Informed by this study, we designed a web-based application as a probe that represents personal data through generative images utilizing Open AI's GPT-4 model and DALL-E 3. We then conducted a 21-day diary study and interviews using the probe with 16 participants to investigate users' in-depth experiences with images generated by AI in everyday lives. Our findings reveal new qualities of experiences in users' engagement with data, highlighting how participants constructed personal meaning from their data through imagination and speculation on AI-generated images. We conclude by discussing the potential and concerns of leveraging image-generative AI for personal data meaning-making. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18853v1-abstract-full').style.display = 'none'; document.getElementById('2502.18853v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">21 pages excluding reference and appendix. Accepted at ACM CHI 2025</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> H.5.0 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.18851">arXiv:2502.18851</a> <span> [<a href="https://arxiv.org/pdf/2502.18851">pdf</a>, <a href="https://arxiv.org/format/2502.18851">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</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"> Marking Code Without Breaking It: Code Watermarking for Detecting LLM-Generated Code </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+J">Jungin Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Shinwoo Park</a>, <a href="/search/cs?searchtype=author&query=Han%2C+Y">Yo-Sub Han</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="2502.18851v1-abstract-short" style="display: inline;"> Code watermarking identifies AI-generated code by embedding patterns into the code during generation. Effective watermarking requires meeting two key conditions: the watermark should be reliably detectable, and the code should retain its original functionality. However, existing methods often modify tokens that are critical for program logic, such as keywords in conditional expressions or operator… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18851v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18851v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18851v1-abstract-full" style="display: none;"> Code watermarking identifies AI-generated code by embedding patterns into the code during generation. Effective watermarking requires meeting two key conditions: the watermark should be reliably detectable, and the code should retain its original functionality. However, existing methods often modify tokens that are critical for program logic, such as keywords in conditional expressions or operators in arithmetic computations. These modifications can cause syntax errors or functional failures, limiting the practical use of watermarking. We present STONE, a method that preserves functional integrity by selectively inserting watermarks only into non-syntax tokens. By excluding tokens essential for code execution, STONE minimizes the risk of functional degradation. In addition, we introduce CWEM, a comprehensive evaluation metric that evaluates watermarking techniques based on correctness, detectability, and naturalness. While correctness and detectability have been widely used, naturalness remains underexplored despite its importance. Unnatural patterns can reveal the presence of a watermark, making it easier for adversaries to remove. We evaluate STONE using CWEM and compare its performance with the state-of-the-art approach. The results show that STONE achieves an average improvement of 7.69% in CWEM across Python, C++, and Java. Our code is available in https://github.com/inistory/STONE-watermarking/. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18851v1-abstract-full').style.display = 'none'; document.getElementById('2502.18851v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.18478">arXiv:2502.18478</a> <span> [<a href="https://arxiv.org/pdf/2502.18478">pdf</a>, <a href="https://arxiv.org/format/2502.18478">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Information Retrieval">cs.IR</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"> Beyond Self-Consistency: Loss-Balanced Perturbation-Based Regularization Improves Industrial-Scale Ads Ranking </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ramazanli%2C+I">Ilqar Ramazanli</a>, <a href="/search/cs?searchtype=author&query=Eghbalzadeh%2C+H">Hamid Eghbalzadeh</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xiaoyi Liu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yang Wang</a>, <a href="/search/cs?searchtype=author&query=Fu%2C+J">Jiaxiang Fu</a>, <a href="/search/cs?searchtype=author&query=Rangadurai%2C+K">Kaushik Rangadurai</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sem Park</a>, <a href="/search/cs?searchtype=author&query=Long%2C+B">Bo Long</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+X">Xue Feng</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="2502.18478v1-abstract-short" style="display: inline;"> Perturbation-based regularization techniques address many challenges in industrial-scale large models, particularly with sparse labels, and emphasize consistency and invariance for perturbation in model predictions. One of the popular regularization techniques has been various forms of self-consistency, which involve making small modifications to input data while preserving contextual information… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18478v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18478v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18478v1-abstract-full" style="display: none;"> Perturbation-based regularization techniques address many challenges in industrial-scale large models, particularly with sparse labels, and emphasize consistency and invariance for perturbation in model predictions. One of the popular regularization techniques has been various forms of self-consistency, which involve making small modifications to input data while preserving contextual information and enforcing similar predictions through auxiliary loss functions. In this work, we explore the first successful application of perturbation-based regularization algorithms in large-scale ads ranking models, and further propose a novel regularization algorithm, namely, Loss-Balanced Small Perturbation Regularization (LSPR) that can be used in potentially any deep learning model. We have successfully demonstrate that both Self-Consistency Regularization approaches (SCR) and LSPR are scalable and can improve ads delivery systems. By conducting industrial-scale experiments, and numerical analysis, we additionally show that our proposed LSPR, performs consistently better compared to SCR, across various groups and signal availability setups. Finally, we report a successful application of the proposed LSPR in a billion-scale industrial ranking system, which to the best of our knowledge, is the first of its kind, and it is specially designed to address the various scalability challenges (e.g, various surfaces, geological locations, clients and so on) as we will mention in this paper. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18478v1-abstract-full').style.display = 'none'; document.getElementById('2502.18478v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.17935">arXiv:2502.17935</a> <span> [<a href="https://arxiv.org/pdf/2502.17935">pdf</a>, <a href="https://arxiv.org/format/2502.17935">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</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/10.1145/3706598.3714226">10.1145/3706598.3714226 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> Less Talk, More Trust: Understanding Players' In-game Assessment of Communication Processes in League of Legends </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lee%2C+J">Juhoon Lee</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Seoyoung Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+Y+S">Yeon Su Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+J">Juho Kim</a>, <a href="/search/cs?searchtype=author&query=Jang%2C+J">Jeong-woo Jang</a>, <a href="/search/cs?searchtype=author&query=Seering%2C+J">Joseph Seering</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="2502.17935v1-abstract-short" style="display: inline;"> In-game team communication in online multiplayer games has shown the potential to foster efficient collaboration and positive social interactions. Yet players often associate communication within ad hoc teams with frustration and wariness. Though previous works have quantitatively analyzed communication patterns at scale, few have identified the motivations of how a player makes in-the-moment comm… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17935v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17935v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17935v1-abstract-full" style="display: none;"> In-game team communication in online multiplayer games has shown the potential to foster efficient collaboration and positive social interactions. Yet players often associate communication within ad hoc teams with frustration and wariness. Though previous works have quantitatively analyzed communication patterns at scale, few have identified the motivations of how a player makes in-the-moment communication decisions. In this paper, we conducted an observation study with 22 League of Legends players by interviewing them during Solo Ranked games on their use of four in-game communication media (chat, pings, emotes, votes). We performed thematic analysis to understand players' in-context assessment and perception of communication attempts. We demonstrate that players evaluate communication opportunities on proximate game states bound by player expectations and norms. Our findings illustrate players' tendency to view communication, regardless of its content, as a precursor to team breakdowns. We build upon these findings to motivate effective player-oriented communication design in online games. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17935v1-abstract-full').style.display = 'none'; document.getElementById('2502.17935v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">17 pages, 2 figures, Accepted by ACM CHI 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.17856">arXiv:2502.17856</a> <span> [<a href="https://arxiv.org/pdf/2502.17856">pdf</a>, <a href="https://arxiv.org/format/2502.17856">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</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/10.1145/3706598.3714218">10.1145/3706598.3714218 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> I Stan Alien Idols and Also the People Behind Them: Understanding How Seams Between Virtual and Real Identities Engage VTuber Fans -- A Case Study of PLAVE </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ahn%2C+D">Dakyeom Ahn</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seora Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+S">Seolhee Lee</a>, <a href="/search/cs?searchtype=author&query=Cho%2C+J">Jieun Cho</a>, <a href="/search/cs?searchtype=author&query=Lim%2C+H">Hajin Lim</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="2502.17856v1-abstract-short" style="display: inline;"> Virtual YouTubers (VTubers) have recently gained popularity as streamers using computer-generated avatars and real-time motion capture to create distinct virtual identities. While prior research has explored how VTubers construct virtual personas and engage audiences, little attention has been given to viewers' reactions when virtual and real identities blur-what we refer to as "seams." To address… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17856v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17856v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17856v1-abstract-full" style="display: none;"> Virtual YouTubers (VTubers) have recently gained popularity as streamers using computer-generated avatars and real-time motion capture to create distinct virtual identities. While prior research has explored how VTubers construct virtual personas and engage audiences, little attention has been given to viewers' reactions when virtual and real identities blur-what we refer to as "seams." To address this gap, we conducted a case study on PLAVE, a popular Korean VTuber Kpop idol group, interviewing 24 of their fans. Our findings identified two main sources of seams: technical glitches and identity collapses, where VTubers act inconsistently with their virtual personas, revealing aspects of their real selves. These seams played a pivotal role in shaping diverse fan engagements, with some valuing authenticity linked to real identities, while others prioritized the coherence of virtual personas. Overall, our findings underscore the importance of seams in shaping viewer experiences. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17856v1-abstract-full').style.display = 'none'; document.getElementById('2502.17856v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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, Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.17749">arXiv:2502.17749</a> <span> [<a href="https://arxiv.org/pdf/2502.17749">pdf</a>, <a href="https://arxiv.org/format/2502.17749">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> </div> </div> <p class="title is-5 mathjax"> Detection of LLM-Paraphrased Code and Identification of the Responsible LLM Using Coding Style Features </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Shinwoo Park</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+H">Hyundong Jin</a>, <a href="/search/cs?searchtype=author&query=Cha%2C+J">Jeong-won Cha</a>, <a href="/search/cs?searchtype=author&query=Han%2C+Y">Yo-Sub Han</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="2502.17749v2-abstract-short" style="display: inline;"> Recent progress in large language models (LLMs) for code generation has raised serious concerns about intellectual property protection. Malicious users can exploit LLMs to produce paraphrased versions of proprietary code that closely resemble the original. While the potential for LLM-assisted code paraphrasing continues to grow, research on detecting it remains limited, underscoring an urgent need… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17749v2-abstract-full').style.display = 'inline'; document.getElementById('2502.17749v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17749v2-abstract-full" style="display: none;"> Recent progress in large language models (LLMs) for code generation has raised serious concerns about intellectual property protection. Malicious users can exploit LLMs to produce paraphrased versions of proprietary code that closely resemble the original. While the potential for LLM-assisted code paraphrasing continues to grow, research on detecting it remains limited, underscoring an urgent need for detection system. We respond to this need by proposing two tasks. The first task is to detect whether code generated by an LLM is a paraphrased version of original human-written code. The second task is to identify which LLM is used to paraphrase the original code. For these tasks, we construct a dataset LPcode consisting of pairs of human-written code and LLM-paraphrased code using various LLMs. We statistically confirm significant differences in the coding styles of human-written and LLM-paraphrased code, particularly in terms of naming consistency, code structure, and readability. Based on these findings, we develop LPcodedec, a detection method that identifies paraphrase relationships between human-written and LLM-generated code, and discover which LLM is used for the paraphrasing. LPcodedec outperforms the best baselines in two tasks, improving F1 scores by 2.64% and 15.17% while achieving speedups of 1,343x and 213x, respectively. Our code and data are available at https://github.com/Shinwoo-Park/detecting_llm_paraphrased_code_via_coding_style_features. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17749v2-abstract-full').style.display = 'none'; document.getElementById('2502.17749v2-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 24 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.16898">arXiv:2502.16898</a> <span> [<a href="https://arxiv.org/pdf/2502.16898">pdf</a>, <a href="https://arxiv.org/format/2502.16898">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"> Variations of Augmented Lagrangian for Robotic Multi-Contact Simulation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jeongmin Lee</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+M">Minji Lee</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sunkyung Park</a>, <a href="/search/cs?searchtype=author&query=Yun%2C+J">Jinhee Yun</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+D">Dongjun 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="2502.16898v1-abstract-short" style="display: inline;"> The multi-contact nonlinear complementarity problem (NCP) is a naturally arising challenge in robotic simulations. Achieving high performance in terms of both accuracy and efficiency remains a significant challenge, particularly in scenarios involving intensive contacts and stiff interactions. In this article, we introduce a new class of multi-contact NCP solvers based on the theory of the Augment… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16898v1-abstract-full').style.display = 'inline'; document.getElementById('2502.16898v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.16898v1-abstract-full" style="display: none;"> The multi-contact nonlinear complementarity problem (NCP) is a naturally arising challenge in robotic simulations. Achieving high performance in terms of both accuracy and efficiency remains a significant challenge, particularly in scenarios involving intensive contacts and stiff interactions. In this article, we introduce a new class of multi-contact NCP solvers based on the theory of the Augmented Lagrangian (AL). We detail how the standard derivation of AL in convex optimization can be adapted to handle multi-contact NCP through the iteration of surrogate problem solutions and the subsequent update of primal-dual variables. Specifically, we present two tailored variations of AL for robotic simulations: the Cascaded Newton-based Augmented Lagrangian (CANAL) and the Subsystem-based Alternating Direction Method of Multipliers (SubADMM). We demonstrate how CANAL can manage multi-contact NCP in an accurate and robust manner, while SubADMM offers superior computational speed, scalability, and parallelizability for high degrees-of-freedom multibody systems with numerous contacts. Our results showcase the effectiveness of the proposed solver framework, illustrating its advantages in various robotic manipulation scenarios. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16898v1-abstract-full').style.display = 'none'; document.getElementById('2502.16898v1-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> 24 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.15422">arXiv:2502.15422</a> <span> [<a href="https://arxiv.org/pdf/2502.15422">pdf</a>, <a href="https://arxiv.org/format/2502.15422">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> <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"> Evaluating Multimodal Generative AI with Korean Educational Standards </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sanghee Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+G">Geewook Kim</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="2502.15422v1-abstract-short" style="display: inline;"> This paper presents the Korean National Educational Test Benchmark (KoNET), a new benchmark designed to evaluate Multimodal Generative AI Systems using Korean national educational tests. KoNET comprises four exams: the Korean Elementary General Educational Development Test (KoEGED), Middle (KoMGED), High (KoHGED), and College Scholastic Ability Test (KoCSAT). These exams are renowned for their rig… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15422v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15422v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15422v1-abstract-full" style="display: none;"> This paper presents the Korean National Educational Test Benchmark (KoNET), a new benchmark designed to evaluate Multimodal Generative AI Systems using Korean national educational tests. KoNET comprises four exams: the Korean Elementary General Educational Development Test (KoEGED), Middle (KoMGED), High (KoHGED), and College Scholastic Ability Test (KoCSAT). These exams are renowned for their rigorous standards and diverse questions, facilitating a comprehensive analysis of AI performance across different educational levels. By focusing on Korean, KoNET provides insights into model performance in less-explored languages. We assess a range of models - open-source, open-access, and closed APIs - by examining difficulties, subject diversity, and human error rates. The code and dataset builder will be made fully open-sourced at https://github.com/naver-ai/KoNET. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15422v1-abstract-full').style.display = 'none'; document.getElementById('2502.15422v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">18 pages; To appear at NAACL 2025 Main Conference (Project page: https://github.com/naver-ai/KoNET )</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.15215">arXiv:2502.15215</a> <span> [<a href="https://arxiv.org/pdf/2502.15215">pdf</a>, <a href="https://arxiv.org/format/2502.15215">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">stat.ML</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="Statistics Theory">math.ST</span> </div> </div> <p class="title is-5 mathjax"> Tensor Product Neural Networks for Functional ANOVA Model </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Seokhun Park</a>, <a href="/search/cs?searchtype=author&query=Kong%2C+I">Insung Kong</a>, <a href="/search/cs?searchtype=author&query=Choi%2C+Y">Yongchan Choi</a>, <a href="/search/cs?searchtype=author&query=Park%2C+C">Chanmoo Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+Y">Yongdai Kim</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="2502.15215v3-abstract-short" style="display: inline;"> Interpretability for machine learning models is becoming more and more important as machine learning models become more complex. The functional ANOVA model, which decomposes a high-dimensional function into a sum of lower dimensional functions (commonly referred to as components), is one of the most popular tools for interpretable AI, and recently, various neural networks have been developed for e… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15215v3-abstract-full').style.display = 'inline'; document.getElementById('2502.15215v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15215v3-abstract-full" style="display: none;"> Interpretability for machine learning models is becoming more and more important as machine learning models become more complex. The functional ANOVA model, which decomposes a high-dimensional function into a sum of lower dimensional functions (commonly referred to as components), is one of the most popular tools for interpretable AI, and recently, various neural networks have been developed for estimating each component in the functional ANOVA model. However, such neural networks are highly unstable when estimating each component since the components themselves are not uniquely defined. That is, there are multiple functional ANOVA decompositions for a given function. In this paper, we propose a novel neural network which guarantees a unique functional ANOVA decomposition and thus is able to estimate each component stably and accurately. We call our proposed neural network ANOVA Tensor Product Neural Network (ANOVA-TPNN) since it is motivated by the tensor product basis expansion. Theoretically, we prove that ANOVA-TPNN can approximate any smooth function well. Empirically, we show that ANOVA-TPNN provide much more stable estimation of each component and thus much more stable interpretation when training data and initial values of the model parameters vary than existing neural networks do. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15215v3-abstract-full').style.display = 'none'; document.getElementById('2502.15215v3-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 21 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">45 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/2502.15054">arXiv:2502.15054</a> <span> [<a href="https://arxiv.org/pdf/2502.15054">pdf</a>, <a href="https://arxiv.org/format/2502.15054">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"> GiGL: Large-Scale Graph Neural Networks at Snapchat </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+T">Tong Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Y">Yozen Liu</a>, <a href="/search/cs?searchtype=author&query=Kolodner%2C+M">Matthew Kolodner</a>, <a href="/search/cs?searchtype=author&query=Montemayor%2C+K">Kyle Montemayor</a>, <a href="/search/cs?searchtype=author&query=Ghazizadeh%2C+E">Elham Ghazizadeh</a>, <a href="/search/cs?searchtype=author&query=Batra%2C+A">Ankit Batra</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+Z">Zihao Fan</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+X">Xiaobin Gao</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+X">Xuan Guo</a>, <a href="/search/cs?searchtype=author&query=Ren%2C+J">Jiwen Ren</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Serim Park</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+P">Peicheng Yu</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+J">Jun Yu</a>, <a href="/search/cs?searchtype=author&query=Vij%2C+S">Shubham Vij</a>, <a href="/search/cs?searchtype=author&query=Shah%2C+N">Neil Shah</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="2502.15054v1-abstract-short" style="display: inline;"> Recent advances in graph machine learning (ML) with the introduction of Graph Neural Networks (GNNs) have led to a widespread interest in applying these approaches to business applications at scale. GNNs enable differentiable end-to-end (E2E) learning of model parameters given graph structure which enables optimization towards popular node, edge (link) and graph-level tasks. While the research inn… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15054v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15054v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15054v1-abstract-full" style="display: none;"> Recent advances in graph machine learning (ML) with the introduction of Graph Neural Networks (GNNs) have led to a widespread interest in applying these approaches to business applications at scale. GNNs enable differentiable end-to-end (E2E) learning of model parameters given graph structure which enables optimization towards popular node, edge (link) and graph-level tasks. While the research innovation in new GNN layers and training strategies has been rapid, industrial adoption and utility of GNNs has lagged considerably due to the unique scale challenges that large-scale graph ML problems create. In this work, we share our approach to training, inference, and utilization of GNNs at Snapchat. To this end, we present GiGL (Gigantic Graph Learning), an open-source library to enable large-scale distributed graph ML to the benefit of researchers, ML engineers, and practitioners. We use GiGL internally at Snapchat to manage the heavy lifting of GNN workflows, including graph data preprocessing from relational DBs, subgraph sampling, distributed training, inference, and orchestration. GiGL is designed to interface cleanly with open-source GNN modeling libraries prominent in academia like PyTorch Geometric (PyG), while handling scaling and productionization challenges that make it easier for internal practitioners to focus on modeling. GiGL is used in multiple production settings, and has powered over 35 launches across multiple business domains in the last 2 years in the contexts of friend recommendation, content recommendation and advertising. This work details high-level design and tools the library provides, scaling properties, case studies in diverse business settings with industry-scale graphs, and several key lessons learned in employing graph ML at scale on large social data. GiGL is open-sourced at https://github.com/snap-research/GiGL. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15054v1-abstract-full').style.display = 'none'; document.getElementById('2502.15054v1-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> 20 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.11948">arXiv:2502.11948</a> <span> [<a href="https://arxiv.org/pdf/2502.11948">pdf</a>, <a href="https://arxiv.org/format/2502.11948">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"> Can Your Uncertainty Scores Detect Hallucinated Entity? </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yeh%2C+M">Min-Hsuan Yeh</a>, <a href="/search/cs?searchtype=author&query=Kamachee%2C+M">Max Kamachee</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seongheon Park</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yixuan 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="2502.11948v1-abstract-short" style="display: inline;"> To mitigate the impact of hallucination nature of LLMs, many studies propose detecting hallucinated generation through uncertainty estimation. However, these approaches predominantly operate at the sentence or paragraph level, failing to pinpoint specific spans or entities responsible for hallucinated content. This lack of granularity is especially problematic for long-form outputs that mix accura… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11948v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11948v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11948v1-abstract-full" style="display: none;"> To mitigate the impact of hallucination nature of LLMs, many studies propose detecting hallucinated generation through uncertainty estimation. However, these approaches predominantly operate at the sentence or paragraph level, failing to pinpoint specific spans or entities responsible for hallucinated content. This lack of granularity is especially problematic for long-form outputs that mix accurate and fabricated information. To address this limitation, we explore entity-level hallucination detection. We propose a new data set, HalluEntity, which annotates hallucination at the entity level. Based on the dataset, we comprehensively evaluate uncertainty-based hallucination detection approaches across 17 modern LLMs. Our experimental results show that uncertainty estimation approaches focusing on individual token probabilities tend to over-predict hallucinations, while context-aware methods show better but still suboptimal performance. Through an in-depth qualitative study, we identify relationships between hallucination tendencies and linguistic properties and highlight important directions for future research. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11948v1-abstract-full').style.display = 'none'; document.getElementById('2502.11948v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.10995">arXiv:2502.10995</a> <span> [<a href="https://arxiv.org/pdf/2502.10995">pdf</a>, <a href="https://arxiv.org/format/2502.10995">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"> Evaluating Large language models on Understanding Korean indirect Speech acts </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Koo%2C+Y">Youngeun Koo</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+J">Jiwoo Lee</a>, <a href="/search/cs?searchtype=author&query=Park%2C+D">Dojun Park</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seohyun Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+S">Sungeun 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="2502.10995v1-abstract-short" style="display: inline;"> To accurately understand the intention of an utterance is crucial in conversational communication. As conversational artificial intelligence models are rapidly being developed and applied in various fields, it is important to evaluate the LLMs' capabilities of understanding the intentions of user's utterance. This study evaluates whether current LLMs can understand the intention of an utterance by… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10995v1-abstract-full').style.display = 'inline'; document.getElementById('2502.10995v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10995v1-abstract-full" style="display: none;"> To accurately understand the intention of an utterance is crucial in conversational communication. As conversational artificial intelligence models are rapidly being developed and applied in various fields, it is important to evaluate the LLMs' capabilities of understanding the intentions of user's utterance. This study evaluates whether current LLMs can understand the intention of an utterance by considering the given conversational context, particularly in cases where the actual intention differs from the surface-leveled, literal intention of the sentence, i.e. indirect speech acts. Our findings reveal that Claude3-Opus outperformed the other competing models, with 71.94% in MCQ and 65% in OEQ, showing a clear advantage. In general, proprietary models exhibited relatively higher performance compared to open-source models. Nevertheless, no LLMs reached the level of human performance. Most LLMs, except for Claude3-Opus, demonstrated significantly lower performance in understanding indirect speech acts compared to direct speech acts, where the intention is explicitly revealed through the utterance. This study not only performs an overall pragmatic evaluation of each LLM's language use through the analysis of OEQ response patterns, but also emphasizes the necessity for further research to improve LLMs' understanding of indirect speech acts for more natural communication with humans. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10995v1-abstract-full').style.display = 'none'; document.getElementById('2502.10995v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 15 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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 (15 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/2502.09955">arXiv:2502.09955</a> <span> [<a href="https://arxiv.org/pdf/2502.09955">pdf</a>, <a href="https://arxiv.org/format/2502.09955">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> </div> </div> <p class="title is-5 mathjax"> Diverse Inference and Verification for Advanced Reasoning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Drori%2C+I">Iddo Drori</a>, <a href="/search/cs?searchtype=author&query=Longhitano%2C+G">Gaston Longhitano</a>, <a href="/search/cs?searchtype=author&query=Mao%2C+M">Mao Mao</a>, <a href="/search/cs?searchtype=author&query=Hyun%2C+S">Seunghwan Hyun</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yuke Zhang</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sungjun Park</a>, <a href="/search/cs?searchtype=author&query=Meeks%2C+Z">Zachary Meeks</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+X">Xin-Yu Zhang</a>, <a href="/search/cs?searchtype=author&query=Segev%2C+B">Ben Segev</a>, <a href="/search/cs?searchtype=author&query=Yong%2C+H">Howard Yong</a>, <a href="/search/cs?searchtype=author&query=Verma%2C+N">Nakul Verma</a>, <a href="/search/cs?searchtype=author&query=Shporer%2C+A">Avi Shporer</a>, <a href="/search/cs?searchtype=author&query=Amit%2C+A">Alon Amit</a>, <a href="/search/cs?searchtype=author&query=Udell%2C+M">Madeleine Udell</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="2502.09955v1-abstract-short" style="display: inline;"> Reasoning LLMs such as OpenAI o1, o3 and DeepSeek R1 have made significant progress in mathematics and coding, yet find challenging advanced tasks such as International Mathematical Olympiad (IMO) combinatorics problems, Abstraction and Reasoning Corpus (ARC) puzzles, and Humanity's Last Exam (HLE) questions. We use a diverse inference approach that combines multiple models and methods at test tim… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09955v1-abstract-full').style.display = 'inline'; document.getElementById('2502.09955v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.09955v1-abstract-full" style="display: none;"> Reasoning LLMs such as OpenAI o1, o3 and DeepSeek R1 have made significant progress in mathematics and coding, yet find challenging advanced tasks such as International Mathematical Olympiad (IMO) combinatorics problems, Abstraction and Reasoning Corpus (ARC) puzzles, and Humanity's Last Exam (HLE) questions. We use a diverse inference approach that combines multiple models and methods at test time. We find that verifying mathematics and code problems, and rejection sampling on other problems is simple and effective. We automatically verify correctness of solutions to IMO problems by Lean, and ARC puzzles by code, and find that best-of-N effectively answers HLE questions. Our approach increases answer accuracy on IMO combinatorics problems from 33.3% to 77.8%, accuracy on HLE questions from 8% to 37%, and solves 80% of ARC puzzles that 948 humans could not and 26.5% of ARC puzzles that o3 high compute does not. Test-time simulations, reinforcement learning, and meta-learning with inference feedback improve generalization by adapting agent graph representations and varying prompts, code, and datasets. Our approach is reliable, robust, and scalable, and in the spirit of reproducible research, we will make it publicly available upon publication. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09955v1-abstract-full').style.display = 'none'; document.getElementById('2502.09955v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">165 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/2502.09065">arXiv:2502.09065</a> <span> [<a href="https://arxiv.org/pdf/2502.09065">pdf</a>, <a href="https://arxiv.org/ps/2502.09065">ps</a>, <a href="https://arxiv.org/format/2502.09065">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Information Theory">cs.IT</span> </div> </div> <p class="title is-5 mathjax"> Lowering the Error Floor of Error Correction Code Transformer </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+T">Taewoo Park</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Seong-Joon Park</a>, <a href="/search/cs?searchtype=author&query=Kwak%2C+H">Hee-Youl Kwak</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sang-Hyo Kim</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+Y">Yongjune Kim</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="2502.09065v1-abstract-short" style="display: inline;"> With the success of transformer architectures across diverse applications, the error correction code transformer (ECCT) has gained significant attention for its superior decoding performance. In spite of its advantages, the error floor phenomenon in ECCT decoding remains unexplored. We present the first investigation of the error floor issue in ECCT and propose a hybrid decoding approach that inte… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09065v1-abstract-full').style.display = 'inline'; document.getElementById('2502.09065v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.09065v1-abstract-full" style="display: none;"> With the success of transformer architectures across diverse applications, the error correction code transformer (ECCT) has gained significant attention for its superior decoding performance. In spite of its advantages, the error floor phenomenon in ECCT decoding remains unexplored. We present the first investigation of the error floor issue in ECCT and propose a hybrid decoding approach that integrates hard decision decoders as pre- and post-decoders with ECCT to effectively lower the error floor. In particular, we introduce a novel loss function for ECCT that considers the dynamics of hybrid decoding algorithm. Training ECCT with the proposed loss function enhances its ability to correct specific error patterns by taking into account its interaction with the auxiliary decoders. Simulation results demonstrate that the proposed hybrid decoder with the novel loss function significantly outperforms the original ECCT in both the waterfall and the error floor regions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09065v1-abstract-full').style.display = 'none'; document.getElementById('2502.09065v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">6 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/2502.08981">arXiv:2502.08981</a> <span> [<a href="https://arxiv.org/pdf/2502.08981">pdf</a>, <a href="https://arxiv.org/format/2502.08981">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</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/10.1145/3706598.3714274">10.1145/3706598.3714274 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> CoCreatAR: Enhancing Authoring of Outdoor Augmented Reality Experiences Through Asymmetric Collaboration </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Numan%2C+N">Nels Numan</a>, <a href="/search/cs?searchtype=author&query=Brostow%2C+G">Gabriel Brostow</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Suhyun Park</a>, <a href="/search/cs?searchtype=author&query=Julier%2C+S">Simon Julier</a>, <a href="/search/cs?searchtype=author&query=Steed%2C+A">Anthony Steed</a>, <a href="/search/cs?searchtype=author&query=Van+Brummelen%2C+J">Jessica Van Brummelen</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="2502.08981v1-abstract-short" style="display: inline;"> Authoring site-specific outdoor augmented reality (AR) experiences requires a nuanced understanding of real-world context to create immersive and relevant content. Existing ex-situ authoring tools typically rely on static 3D models to represent spatial information. However, in our formative study (n=25), we identified key limitations of this approach: models are often outdated, incomplete, or insu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08981v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08981v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08981v1-abstract-full" style="display: none;"> Authoring site-specific outdoor augmented reality (AR) experiences requires a nuanced understanding of real-world context to create immersive and relevant content. Existing ex-situ authoring tools typically rely on static 3D models to represent spatial information. However, in our formative study (n=25), we identified key limitations of this approach: models are often outdated, incomplete, or insufficient for capturing critical factors such as safety considerations, user flow, and dynamic environmental changes. These issues necessitate frequent on-site visits and additional iterations, making the authoring process more time-consuming and resource-intensive. To mitigate these challenges, we introduce CoCreatAR, an asymmetric collaborative mixed reality authoring system that integrates the flexibility of ex-situ workflows with the immediate contextual awareness of in-situ authoring. We conducted an exploratory study (n=32) comparing CoCreatAR to an asynchronous workflow baseline, finding that it enhances engagement, creativity, and confidence in the authored output while also providing preliminary insights into its impact on task load. We conclude by discussing the implications of our findings for integrating real-world context into site-specific AR authoring systems. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08981v1-abstract-full').style.display = 'none'; document.getElementById('2502.08981v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">22 pages, 12 figures, ACM CHI 2025</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.08668">arXiv:2502.08668</a> <span> [<a href="https://arxiv.org/pdf/2502.08668">pdf</a>, <a href="https://arxiv.org/format/2502.08668">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"> Style Extraction on Text Embeddings Using VAE and Parallel Dataset </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kong%2C+I">InJin Kong</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+S">Shinyee Kang</a>, <a href="/search/cs?searchtype=author&query=Park%2C+Y">Yuna Park</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sooyong Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sanghyun Park</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="2502.08668v1-abstract-short" style="display: inline;"> This study investigates the stylistic differences among various Bible translations using a Variational Autoencoder (VAE) model. By embedding textual data into high-dimensional vectors, the study aims to detect and analyze stylistic variations between translations, with a specific focus on distinguishing the American Standard Version (ASV) from other translations. The results demonstrate that each… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08668v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08668v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08668v1-abstract-full" style="display: none;"> This study investigates the stylistic differences among various Bible translations using a Variational Autoencoder (VAE) model. By embedding textual data into high-dimensional vectors, the study aims to detect and analyze stylistic variations between translations, with a specific focus on distinguishing the American Standard Version (ASV) from other translations. The results demonstrate that each translation exhibits a unique stylistic distribution, which can be effectively identified using the VAE model. These findings suggest that the VAE model is proficient in capturing and differentiating textual styles, although it is primarily optimized for distinguishing a single style. The study highlights the model's potential for broader applications in AI-based text generation and stylistic analysis, while also acknowledging the need for further model refinement to address the complexity of multi-dimensional stylistic relationships. Future research could extend this methodology to other text domains, offering deeper insights into the stylistic features embedded within various types of textual data. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08668v1-abstract-full').style.display = 'none'; document.getElementById('2502.08668v1-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> 11 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">9 pages, 6 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/2502.08167">arXiv:2502.08167</a> <span> [<a href="https://arxiv.org/pdf/2502.08167">pdf</a>, <a href="https://arxiv.org/format/2502.08167">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"> DNNs May Determine Major Properties of Their Outputs Early, with Timing Possibly Driven by Bias </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Song Park</a>, <a href="/search/cs?searchtype=author&query=Chun%2C+S">Sanghyuk Chun</a>, <a href="/search/cs?searchtype=author&query=Heo%2C+B">Byeongho Heo</a>, <a href="/search/cs?searchtype=author&query=Han%2C+D">Dongyoon Han</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="2502.08167v1-abstract-short" style="display: inline;"> This paper argues that deep neural networks (DNNs) mostly determine their outputs during the early stages of inference, where biases inherent in the model play a crucial role in shaping this process. We draw a parallel between this phenomenon and human decision-making, which often relies on fast, intuitive heuristics. Using diffusion models (DMs) as a case study, we demonstrate that DNNs often mak… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08167v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08167v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08167v1-abstract-full" style="display: none;"> This paper argues that deep neural networks (DNNs) mostly determine their outputs during the early stages of inference, where biases inherent in the model play a crucial role in shaping this process. We draw a parallel between this phenomenon and human decision-making, which often relies on fast, intuitive heuristics. Using diffusion models (DMs) as a case study, we demonstrate that DNNs often make early-stage decision-making influenced by the type and extent of bias in their design and training. Our findings offer a new perspective on bias mitigation, efficient inference, and the interpretation of machine learning systems. By identifying the temporal dynamics of decision-making in DNNs, this paper aims to inspire further discussion and research within the machine learning community. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08167v1-abstract-full').style.display = 'none'; document.getElementById('2502.08167v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">First two authors contributed equally</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.07836">arXiv:2502.07836</a> <span> [<a href="https://arxiv.org/pdf/2502.07836">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Advancing Precision Oncology Through Modeling of Longitudinal and Multimodal Data </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhuang%2C+L">Luoting Zhuang</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S+H">Stephen H. Park</a>, <a href="/search/cs?searchtype=author&query=Skates%2C+S+J">Steven J. Skates</a>, <a href="/search/cs?searchtype=author&query=Prosper%2C+A+E">Ashley E. Prosper</a>, <a href="/search/cs?searchtype=author&query=Aberle%2C+D+R">Denise R. Aberle</a>, <a href="/search/cs?searchtype=author&query=Hsu%2C+W">William Hsu</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="2502.07836v1-abstract-short" style="display: inline;"> Cancer evolves continuously over time through a complex interplay of genetic, epigenetic, microenvironmental, and phenotypic changes. This dynamic behavior drives uncontrolled cell growth, metastasis, immune evasion, and therapy resistance, posing challenges for effective monitoring and treatment. However, today's data-driven research in oncology has primarily focused on cross-sectional analysis u… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.07836v1-abstract-full').style.display = 'inline'; document.getElementById('2502.07836v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.07836v1-abstract-full" style="display: none;"> Cancer evolves continuously over time through a complex interplay of genetic, epigenetic, microenvironmental, and phenotypic changes. This dynamic behavior drives uncontrolled cell growth, metastasis, immune evasion, and therapy resistance, posing challenges for effective monitoring and treatment. However, today's data-driven research in oncology has primarily focused on cross-sectional analysis using data from a single modality, limiting the ability to fully characterize and interpret the disease's dynamic heterogeneity. Advances in multiscale data collection and computational methods now enable the discovery of longitudinal multimodal biomarkers for precision oncology. Longitudinal data reveal patterns of disease progression and treatment response that are not evident from single-timepoint data, enabling timely abnormality detection and dynamic treatment adaptation. Multimodal data integration offers complementary information from diverse sources for more precise risk assessment and targeting of cancer therapy. In this review, we survey methods of longitudinal and multimodal modeling, highlighting their synergy in providing multifaceted insights for personalized care tailored to the unique characteristics of a patient's cancer. We summarize the current challenges and future directions of longitudinal multimodal analysis in advancing precision oncology. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.07836v1-abstract-full').style.display = 'none'; document.getElementById('2502.07836v1-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> 10 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">This work has been submitted to the IEEE RBME for potential publication</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06834">arXiv:2502.06834</a> <span> [<a href="https://arxiv.org/pdf/2502.06834">pdf</a>, <a href="https://arxiv.org/format/2502.06834">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"> A Unified Knowledge-Distillation and Semi-Supervised Learning Framework to Improve Industrial Ads Delivery Systems </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Eghbalzadeh%2C+H">Hamid Eghbalzadeh</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yang Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+R">Rui Li</a>, <a href="/search/cs?searchtype=author&query=Mo%2C+Y">Yuji Mo</a>, <a href="/search/cs?searchtype=author&query=Ding%2C+Q">Qin Ding</a>, <a href="/search/cs?searchtype=author&query=Fu%2C+J">Jiaxiang Fu</a>, <a href="/search/cs?searchtype=author&query=Dai%2C+L">Liang Dai</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+S">Shuo Gu</a>, <a href="/search/cs?searchtype=author&query=Noorshams%2C+N">Nima Noorshams</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sem Park</a>, <a href="/search/cs?searchtype=author&query=Long%2C+B">Bo Long</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+X">Xue Feng</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="2502.06834v1-abstract-short" style="display: inline;"> Industrial ads ranking systems conventionally rely on labeled impression data, which leads to challenges such as overfitting, slower incremental gain from model scaling, and biases due to discrepancies between training and serving data. To overcome these issues, we propose a Unified framework for Knowledge-Distillation and Semi-supervised Learning (UKDSL) for ads ranking, empowering the training o… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06834v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06834v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06834v1-abstract-full" style="display: none;"> Industrial ads ranking systems conventionally rely on labeled impression data, which leads to challenges such as overfitting, slower incremental gain from model scaling, and biases due to discrepancies between training and serving data. To overcome these issues, we propose a Unified framework for Knowledge-Distillation and Semi-supervised Learning (UKDSL) for ads ranking, empowering the training of models on a significantly larger and more diverse datasets, thereby reducing overfitting and mitigating training-serving data discrepancies. We provide detailed formal analysis and numerical simulations on the inherent miscalibration and prediction bias of multi-stage ranking systems, and show empirical evidence of the proposed framework's capability to mitigate those. Compared to prior work, UKDSL can enable models to learn from a much larger set of unlabeled data, hence, improving the performance while being computationally efficient. Finally, we report the successful deployment of UKDSL in an industrial setting across various ranking models, serving users at multi-billion scale, across various surfaces, geological locations, clients, and optimize for various events, which to the best of our knowledge is the first of its kind in terms of the scale and efficiency at which it operates. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06834v1-abstract-full').style.display = 'none'; document.getElementById('2502.06834v1-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06352">arXiv:2502.06352</a> <span> [<a href="https://arxiv.org/pdf/2502.06352">pdf</a>, <a href="https://arxiv.org/format/2502.06352">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"> LANTERN++: Enhancing Relaxed Speculative Decoding with Static Tree Drafting for Visual Auto-regressive Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Park%2C+S">Sihwan Park</a>, <a href="/search/cs?searchtype=author&query=Jang%2C+D">Doohyuk Jang</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sungyub Kim</a>, <a href="/search/cs?searchtype=author&query=Kundu%2C+S">Souvik Kundu</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+E">Eunho 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="2502.06352v2-abstract-short" style="display: inline;"> Speculative decoding has been widely used to accelerate auto-regressive (AR) text generation. However, its effectiveness for visual AR models remains limited due to token selection ambiguity, where multiple tokens share similarly low probabilities and thus reduce acceptance rates. Recently, relaxed speculative decoding with dynamic tree drafting was proposed to mitigate this ambiguity, demonstrati… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06352v2-abstract-full').style.display = 'inline'; document.getElementById('2502.06352v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06352v2-abstract-full" style="display: none;"> Speculative decoding has been widely used to accelerate auto-regressive (AR) text generation. However, its effectiveness for visual AR models remains limited due to token selection ambiguity, where multiple tokens share similarly low probabilities and thus reduce acceptance rates. Recently, relaxed speculative decoding with dynamic tree drafting was proposed to mitigate this ambiguity, demonstrating promising results in accelerating visual AR models. However, we observe that token selection ambiguity still negatively affects dynamic tree drafting, resulting in shallow draft trees and limited acceleration. To overcome this issue, we introduce LANTERN++, a refined framework that integrates static tree drafting with a tailored relaxed acceptance condition, allowing drafts to be selected independently of low-confidence predictions. This enables the acceptance of deeper sequences, improving decoding efficiency while preserving image quality. Extensive experiments on state-of-the-art visual AR models demonstrate that LANTERN++ significantly accelerates inference, achieving up to $\mathbf{\times 2.56}$ speedup over standard AR decoding while maintaining high image quality. The code is publicly available at https://github.com/jadohu/LANTERN. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06352v2-abstract-full').style.display = 'none'; document.getElementById('2502.06352v2-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 March, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 10 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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">ICLR 2025 Workshop at SCOPE (Oral), 16 pages, 5 figures, short paper (6 pages exclude reference and appendix)</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06142">arXiv:2502.06142</a> <span> [<a href="https://arxiv.org/pdf/2502.06142">pdf</a>, <a href="https://arxiv.org/format/2502.06142">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">stat.ML</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"> Linear Bandits with Partially Observable Features </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+W">Wonyoung Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sungwoo Park</a>, <a href="/search/cs?searchtype=author&query=Iyengar%2C+G">Garud Iyengar</a>, <a href="/search/cs?searchtype=author&query=Zeevi%2C+A">Assaf Zeevi</a>, <a href="/search/cs?searchtype=author&query=Oh%2C+M">Min-hwan Oh</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="2502.06142v1-abstract-short" style="display: inline;"> We introduce a novel linear bandit problem with partially observable features, resulting in partial reward information and spurious estimates. Without proper address for latent part, regret possibly grows linearly in decision horizon $T$, as their influence on rewards are unknown. To tackle this, we propose a novel analysis to handle the latent features and an algorithm that achieves sublinear reg… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06142v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06142v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06142v1-abstract-full" style="display: none;"> We introduce a novel linear bandit problem with partially observable features, resulting in partial reward information and spurious estimates. Without proper address for latent part, regret possibly grows linearly in decision horizon $T$, as their influence on rewards are unknown. To tackle this, we propose a novel analysis to handle the latent features and an algorithm that achieves sublinear regret. The core of our algorithm involves (i) augmenting basis vectors orthogonal to the observed feature space, and (ii) introducing an efficient doubly robust estimator. Our approach achieves a regret bound of $\tilde{O}(\sqrt{(d + d_h)T})$, where $d$ is the dimension of observed features, and $d_h$ is the unknown dimension of the subspace of the unobserved features. Notably, our algorithm requires no prior knowledge of the unobserved feature space, which may expand as more features become hidden. Numerical experiments confirm that our algorithm outperforms both non-contextual multi-armed bandits and linear bandit algorithms depending solely on observed features. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06142v1-abstract-full').style.display = 'none'; document.getElementById('2502.06142v1-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> 9 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04615">arXiv:2502.04615</a> <span> [<a href="https://arxiv.org/pdf/2502.04615">pdf</a>, <a href="https://arxiv.org/format/2502.04615">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 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/10.1145/3681756.3697973">10.1145/3681756.3697973 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> Neural Clustering for Prefractured Mesh Generation in Real-time Object Destruction </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kim%2C+S">Seunghwan Kim</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sunha Park</a>, <a href="/search/cs?searchtype=author&query=Lee%2C+S">Seungkyu 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="2502.04615v1-abstract-short" style="display: inline;"> Prefracture method is a practical implementation for real-time object destruction that is hardly achievable within performance constraints, but can produce unrealistic results due to its heuristic nature. To mitigate it, we approach the clustering of prefractured mesh generation as an unordered segmentation on point cloud data, and propose leveraging the deep neural network trained on a physics-ba… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04615v1-abstract-full').style.display = 'inline'; document.getElementById('2502.04615v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04615v1-abstract-full" style="display: none;"> Prefracture method is a practical implementation for real-time object destruction that is hardly achievable within performance constraints, but can produce unrealistic results due to its heuristic nature. To mitigate it, we approach the clustering of prefractured mesh generation as an unordered segmentation on point cloud data, and propose leveraging the deep neural network trained on a physics-based dataset. Our novel paradigm successfully predicts the structural weakness of object that have been limited, exhibiting ready-to-use results with remarkable quality. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04615v1-abstract-full').style.display = 'none'; document.getElementById('2502.04615v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 6 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04538">arXiv:2502.04538</a> <span> [<a href="https://arxiv.org/pdf/2502.04538">pdf</a>, <a href="https://arxiv.org/format/2502.04538">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</span> </div> </div> <p class="title is-5 mathjax"> SoK: "Interoperability vs Security" Arguments: A Technical Framework </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Landis%2C+D">Daji Landis</a>, <a href="/search/cs?searchtype=author&query=Bietti%2C+E">Elettra Bietti</a>, <a href="/search/cs?searchtype=author&query=Park%2C+S">Sunoo Park</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="2502.04538v1-abstract-short" style="display: inline;"> Concerns about big tech's monopoly power have featured prominently in recent media and policy discourse, and regulators across the US, the EU, and beyond have ramped up efforts to promote healthier competition in the market. One of the favored approaches is to require certain kinds of interoperation between platforms, to mitigate the current concentration of power in the biggest companies. Unsurpr… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04538v1-abstract-full').style.display = 'inline'; document.getElementById('2502.04538v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04538v1-abstract-full" style="display: none;"> Concerns about big tech's monopoly power have featured prominently in recent media and policy discourse, and regulators across the US, the EU, and beyond have ramped up efforts to promote healthier competition in the market. One of the favored approaches is to require certain kinds of interoperation between platforms, to mitigate the current concentration of power in the biggest companies. Unsurprisingly, interoperability initiatives have generally been met with vocal resistance by big tech companies. Perhaps more surprisingly, a significant part of that pushback has been in the name of security -- that is, arguing against interoperation on the basis that it will undermine security. We conduct a detailed examination of "security vs. interoperability" arguments in the context of recent antitrust proceedings in the US and the EU. First, we propose a taxonomy of such arguments. Second, we provide several detailed case studies, which illustrate our taxonomy's utility in disentangling where security and interoperability are and are not in tension, where securing interoperable systems presents novel engineering challenges, and where "security arguments" against interoperability are really more about anti-competitive behavior than security. Third, we undertake a comparative analysis that highlights key considerations around the interplay of economic incentives, market power, and security across diverse contexts where security and interoperability may appear to be in tension. We believe systematically distinguishing cases and patterns within our taxonomy and analytical framework can be a valuable analytical tool for experts and non-experts alike in today's fast-paced regulatory landscape. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04538v1-abstract-full').style.display = 'none'; document.getElementById('2502.04538v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 6 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </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=Park%2C+S&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Park%2C+S&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