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 2,651 results for author: <span class="mathjax">Li, B</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=Li%2C+B">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="Li, B"> </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=Li%2C+B&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="Li, B"> <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=Li%2C+B&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&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/2502.18300">arXiv:2502.18300</a> <span> [<a href="https://arxiv.org/pdf/2502.18300">pdf</a>, <a href="https://arxiv.org/format/2502.18300">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="Machine Learning">stat.ML</span> </div> </div> <p class="title is-5 mathjax"> Bayesian Computation in Deep Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+W">Wenlong Chen</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bolian Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+R">Ruqi Zhang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yingzhen 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="2502.18300v1-abstract-short" style="display: inline;"> This review paper is intended for the 2nd edition of the Handbook of Markov chain Monte Carlo.We provide an introduction to approximate inference techniques as Bayesian computation methods applied to deep learning models. We organize the chapter by presenting popular computational methods for (1) Bayesian neural networks and (2) deep generative models, explaining their unique challenges in posteri… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18300v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18300v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18300v1-abstract-full" style="display: none;"> This review paper is intended for the 2nd edition of the Handbook of Markov chain Monte Carlo.We provide an introduction to approximate inference techniques as Bayesian computation methods applied to deep learning models. We organize the chapter by presenting popular computational methods for (1) Bayesian neural networks and (2) deep generative models, explaining their unique challenges in posterior inference as well as the solutions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18300v1-abstract-full').style.display = 'none'; document.getElementById('2502.18300v1-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">43 pages, 7 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/2502.18185">arXiv:2502.18185</a> <span> [<a href="https://arxiv.org/pdf/2502.18185">pdf</a>, <a href="https://arxiv.org/format/2502.18185">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Image and Video Processing">eess.IV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> VesselSAM: Leveraging SAM for Aortic Vessel Segmentation with LoRA and Atrous Attention </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Iltaf%2C+A">Adnan Iltaf</a>, <a href="/search/cs?searchtype=author&query=Ahmed%2C+R+M">Rayan Merghani Ahmed</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin Li</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+S">Shoujun Zhou</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="2502.18185v1-abstract-short" style="display: inline;"> Medical image segmentation is crucial for clinical diagnosis and treatment planning, particularly for complex anatomical structures like vessels. In this work, we propose VesselSAM, a modified version of the Segmentation Anything Model (SAM), specifically designed for aortic vessel segmentation. VesselSAM incorporates AtrousLoRA, a novel module that combines Atrous Attention with Low-Rank Adaptati… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18185v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18185v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18185v1-abstract-full" style="display: none;"> Medical image segmentation is crucial for clinical diagnosis and treatment planning, particularly for complex anatomical structures like vessels. In this work, we propose VesselSAM, a modified version of the Segmentation Anything Model (SAM), specifically designed for aortic vessel segmentation. VesselSAM incorporates AtrousLoRA, a novel module that combines Atrous Attention with Low-Rank Adaptation (LoRA), to improve segmentation performance. Atrous Attention enables the model to capture multi-scale contextual information, preserving both fine local details and broader global context. At the same time, LoRA facilitates efficient fine-tuning of the frozen SAM image encoder, reducing the number of trainable parameters and ensuring computational efficiency. We evaluate VesselSAM on two challenging datasets: the Aortic Vessel Tree (AVT) dataset and the Type-B Aortic Dissection (TBAD) dataset. VesselSAM achieves state-of-the-art performance with DSC scores of 93.50\%, 93.25\%, 93.02\%, and 93.26\% across multiple medical centers. Our results demonstrate that VesselSAM delivers high segmentation accuracy while significantly reducing computational overhead compared to existing large-scale models. This development paves the way for enhanced AI-based aortic vessel segmentation in clinical environments. The code and models will be released at https://github.com/Adnan-CAS/AtrousLora. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18185v1-abstract-full').style.display = 'none'; document.getElementById('2502.18185v1-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">Submitted to IEEE JBHI</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.17709">arXiv:2502.17709</a> <span> [<a href="https://arxiv.org/pdf/2502.17709">pdf</a>, <a href="https://arxiv.org/format/2502.17709">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="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Multimedia">cs.MM</span> </div> </div> <p class="title is-5 mathjax"> Contrastive Visual Data Augmentation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhou%2C+Y">Yu Zhou</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bingxuan Li</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+M">Mohan Tang</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+X">Xiaomeng Jin</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+T">Te-Lin Wu</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+K">Kuan-Hao Huang</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+H">Heng Ji</a>, <a href="/search/cs?searchtype=author&query=Chang%2C+K">Kai-Wei Chang</a>, <a href="/search/cs?searchtype=author&query=Peng%2C+N">Nanyun Peng</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="2502.17709v1-abstract-short" style="display: inline;"> Large multimodal models (LMMs) often struggle to recognize novel concepts, as they rely on pre-trained knowledge and have limited ability to capture subtle visual details. Domain-specific knowledge gaps in training also make them prone to confusing visually similar, commonly misrepresented, or low-resource concepts. To help LMMs better align nuanced visual features with language, improving their a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17709v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17709v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17709v1-abstract-full" style="display: none;"> Large multimodal models (LMMs) often struggle to recognize novel concepts, as they rely on pre-trained knowledge and have limited ability to capture subtle visual details. Domain-specific knowledge gaps in training also make them prone to confusing visually similar, commonly misrepresented, or low-resource concepts. To help LMMs better align nuanced visual features with language, improving their ability to recognize and reason about novel or rare concepts, we propose a Contrastive visual Data Augmentation (CoDA) strategy. CoDA extracts key contrastive textual and visual features of target concepts against the known concepts they are misrecognized as, and then uses multimodal generative models to produce targeted synthetic data. Automatic filtering of extracted features and augmented images is implemented to guarantee their quality, as verified by human annotators. We show the effectiveness and efficiency of CoDA on low-resource concept and diverse scene recognition datasets including INaturalist and SUN. We additionally collect NovelSpecies, a benchmark dataset consisting of newly discovered animal species that are guaranteed to be unseen by LMMs. LLaVA-1.6 1-shot updating results on these three datasets show CoDA significantly improves SOTA visual data augmentation strategies by 12.3% (NovelSpecies), 5.1% (SUN), and 6.0% (iNat) absolute gains in accuracy. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17709v1-abstract-full').style.display = 'none'; document.getElementById('2502.17709v1-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/2502.17651">arXiv:2502.17651</a> <span> [<a href="https://arxiv.org/pdf/2502.17651">pdf</a>, <a href="https://arxiv.org/format/2502.17651">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="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> METAL: A Multi-Agent Framework for Chart Generation with Test-Time Scaling </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Bingxuan Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yiwei Wang</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+J">Jiuxiang Gu</a>, <a href="/search/cs?searchtype=author&query=Chang%2C+K">Kai-Wei Chang</a>, <a href="/search/cs?searchtype=author&query=Peng%2C+N">Nanyun Peng</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="2502.17651v1-abstract-short" style="display: inline;"> Chart generation aims to generate code to produce charts satisfying the desired visual properties, e.g., texts, layout, color, and type. It has great potential to empower the automatic professional report generation in financial analysis, research presentation, education, and healthcare. In this work, we build a vision-language model (VLM) based multi-agent framework for effective automatic chart… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17651v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17651v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17651v1-abstract-full" style="display: none;"> Chart generation aims to generate code to produce charts satisfying the desired visual properties, e.g., texts, layout, color, and type. It has great potential to empower the automatic professional report generation in financial analysis, research presentation, education, and healthcare. In this work, we build a vision-language model (VLM) based multi-agent framework for effective automatic chart generation. Generating high-quality charts requires both strong visual design skills and precise coding capabilities that embed the desired visual properties into code. Such a complex multi-modal reasoning process is difficult for direct prompting of VLMs. To resolve these challenges, we propose METAL, a multi-agent framework that decomposes the task of chart generation into the iterative collaboration among specialized agents. METAL achieves 5.2% improvement in accuracy over the current best result in the chart generation task. The METAL framework exhibits the phenomenon of test-time scaling: its performance increases monotonically as the logarithmic computational budget grows from 512 to 8192 tokens. In addition, we find that separating different modalities during the critique process of METAL boosts the self-correction capability of VLMs in the multimodal context. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17651v1-abstract-full').style.display = 'none'; document.getElementById('2502.17651v1-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/2502.17535">arXiv:2502.17535</a> <span> [<a href="https://arxiv.org/pdf/2502.17535">pdf</a>, <a href="https://arxiv.org/format/2502.17535">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Formal Languages and Automata Theory">cs.FL</span> </div> </div> <p class="title is-5 mathjax"> The Lottery LLM Hypothesis, Rethinking What Abilities Should LLM Compression Preserve? </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tang%2C+Z">Zhenheng Tang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xiang Liu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Q">Qian Wang</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+P">Peijie Dong</a>, <a href="/search/cs?searchtype=author&query=He%2C+B">Bingsheng He</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+X">Xiaowen Chu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo 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="2502.17535v1-abstract-short" style="display: inline;"> Motivated by reducing the computational and storage costs of LLMs, model compression and KV cache compression have attracted much attention from researchers. However, current methods predominantly emphasize maintaining the performance of compressed LLMs, as measured by perplexity or simple accuracy on tasks of common sense knowledge QA and basic arithmetic reasoning. In this blog, we present a bri… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17535v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17535v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17535v1-abstract-full" style="display: none;"> Motivated by reducing the computational and storage costs of LLMs, model compression and KV cache compression have attracted much attention from researchers. However, current methods predominantly emphasize maintaining the performance of compressed LLMs, as measured by perplexity or simple accuracy on tasks of common sense knowledge QA and basic arithmetic reasoning. In this blog, we present a brief review of recent advancements in LLMs related to retrieval-augmented generation, multi-step reasoning, external tools, and computational expressivity, all of which substantially enhance LLM performance. Then, we propose a lottery LLM hypothesis suggesting that for a given LLM and task, there exists a smaller lottery LLM capable of producing the same performance as the original LLM with the assistance of multi-step reasoning and external tools. Based on the review of current progress in LLMs, we discuss and summarize the essential capabilities that the lottery LLM and KV cache compression must possess, which are currently overlooked in existing methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17535v1-abstract-full').style.display = 'none'; document.getElementById('2502.17535v1-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/2502.17248">arXiv:2502.17248</a> <span> [<a href="https://arxiv.org/pdf/2502.17248">pdf</a>, <a href="https://arxiv.org/format/2502.17248">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Databases">cs.DB</span> </div> </div> <p class="title is-5 mathjax"> Alpha-SQL: Zero-Shot Text-to-SQL using Monte Carlo Tree Search </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Boyan Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jiayi Zhang</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+J">Ju Fan</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+Y">Yanwei Xu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+C">Chong Chen</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+N">Nan Tang</a>, <a href="/search/cs?searchtype=author&query=Luo%2C+Y">Yuyu Luo</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="2502.17248v1-abstract-short" style="display: inline;"> Text-to-SQL, which enables natural language interaction with databases, serves as a pivotal method across diverse industries. With new, more powerful large language models (LLMs) emerging every few months, fine-tuning has become incredibly costly, labor-intensive, and error-prone. As an alternative, zero-shot Text-to-SQL, which leverages the growing knowledge and reasoning capabilities encoded in… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17248v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17248v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17248v1-abstract-full" style="display: none;"> Text-to-SQL, which enables natural language interaction with databases, serves as a pivotal method across diverse industries. With new, more powerful large language models (LLMs) emerging every few months, fine-tuning has become incredibly costly, labor-intensive, and error-prone. As an alternative, zero-shot Text-to-SQL, which leverages the growing knowledge and reasoning capabilities encoded in LLMs without task-specific fine-tuning, presents a promising and more challenging direction. To address this challenge, we propose Alpha-SQL, a novel approach that leverages a Monte Carlo Tree Search (MCTS) framework to iteratively infer SQL construction actions based on partial SQL query states. To enhance the framework's reasoning capabilities, we introduce LLM-as-Action-Model to dynamically generate SQL construction actions during the MCTS process, steering the search toward more promising SQL queries. Moreover, Alpha-SQL employs a self-supervised reward function to evaluate the quality of candidate SQL queries, ensuring more accurate and efficient query generation. Experimental results show that Alpha-SQL achieves 69.7% execution accuracy on the BIRD development set, using a 32B open-source LLM without fine-tuning. Alpha-SQL outperforms the best previous zero-shot approach based on GPT-4o by 2.5% on the BIRD development set. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17248v1-abstract-full').style.display = 'none'; document.getElementById('2502.17248v1-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/2502.17006">arXiv:2502.17006</a> <span> [<a href="https://arxiv.org/pdf/2502.17006">pdf</a>, <a href="https://arxiv.org/format/2502.17006">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Hardware Architecture">cs.AR</span> </div> </div> <p class="title is-5 mathjax"> Be CIM or Be Memory: A Dual-mode-aware DNN Compiler for CIM Accelerators </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhao%2C+S">Shixin Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yuming Li</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bing Li</a>, <a href="/search/cs?searchtype=author&query=He%2C+Y">Yintao He</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+M">Mengdi Wang</a>, <a href="/search/cs?searchtype=author&query=Han%2C+Y">Yinhe Han</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Ying 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="2502.17006v1-abstract-short" style="display: inline;"> Computing-in-memory (CIM) architectures demonstrate superior performance over traditional architectures. To unleash the potential of CIM accelerators, many compilation methods have been proposed, focusing on application scheduling optimization specific to CIM. However, existing compilation methods often overlook CIM's capability to switch dynamically between compute and memory modes, which is cruc… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17006v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17006v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17006v1-abstract-full" style="display: none;"> Computing-in-memory (CIM) architectures demonstrate superior performance over traditional architectures. To unleash the potential of CIM accelerators, many compilation methods have been proposed, focusing on application scheduling optimization specific to CIM. However, existing compilation methods often overlook CIM's capability to switch dynamically between compute and memory modes, which is crucial for accommodating the diverse memory and computational needs of real-world deep neural network architectures, especially the emerging large language models. To fill this gap, we introduce CMSwitch, a novel compiler to optimize resource allocation for CIM accelerators with adaptive mode-switching capabilities, thereby enhancing the performance of DNN applications. Specifically, our approach integrates the compute-memory mode switch into the CIM compilation optimization space by introducing a new hardware abstraction attribute. Then, we propose a novel compilation optimization pass that identifies the optimal network segment and the corresponding mode resource allocations using dynamic programming and mixed-integer programming. CMSwitch uses the tailored meta-operator to express the compilation result in a generalized manner. Evaluation results demonstrate that CMSwitch achieves an average speedup of 1.31$\times$ compared to existing SOTA CIM compilation works, highlighting CMSwitch's effectiveness in fully exploiting the potential of CIM processors for a wide range of real-world DNN applications. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17006v1-abstract-full').style.display = 'none'; document.getElementById('2502.17006v1-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> <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, 18 figures, accepted to ASPLOS 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/2502.16963">arXiv:2502.16963</a> <span> [<a href="https://arxiv.org/pdf/2502.16963">pdf</a>, <a href="https://arxiv.org/format/2502.16963">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Hardware Architecture">cs.AR</span> </div> </div> <p class="title is-5 mathjax"> Make LLM Inference Affordable to Everyone: Augmenting GPU Memory with NDP-DIMM </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Liu%2C+L">Lian Liu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+S">Shixin Zhao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bing Li</a>, <a href="/search/cs?searchtype=author&query=Ren%2C+H">Haimeng Ren</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+Z">Zhaohui Xu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+M">Mengdi Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xiaowei Li</a>, <a href="/search/cs?searchtype=author&query=Han%2C+Y">Yinhe Han</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Ying 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="2502.16963v1-abstract-short" style="display: inline;"> The billion-scale Large Language Models (LLMs) need deployment on expensive server-grade GPUs with large-storage HBMs and abundant computation capability. As LLM-assisted services become popular, achieving cost-effective LLM inference on budget-friendly hardware becomes the trend. Extensive researches relocate LLM parameters from expensive GPUs to host memory. However, the restricted bandwidth bet… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16963v1-abstract-full').style.display = 'inline'; document.getElementById('2502.16963v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.16963v1-abstract-full" style="display: none;"> The billion-scale Large Language Models (LLMs) need deployment on expensive server-grade GPUs with large-storage HBMs and abundant computation capability. As LLM-assisted services become popular, achieving cost-effective LLM inference on budget-friendly hardware becomes the trend. Extensive researches relocate LLM parameters from expensive GPUs to host memory. However, the restricted bandwidth between the host and GPU memory limits the inference performance. This work introduces Hermes, a budget-friendly system that leverages the near-data processing (NDP) within commodity DRAM DIMMs to enhance the performance of a single consumer-grade GPU, achieving efficient LLM inference. The inherent activation sparsity in LLMs naturally divides weight parameters into two categories, termed ``hot" and ``cold" neurons, respectively. Hot neurons, which consist of only approximately 20\% of all weight parameters, account for 80\% of the total computational load, while cold neurons make up the other 80\% of parameters but are responsible for just 20\% of the computational load. Therefore, we propose a heterogeneous computing strategy: mapping hot neurons to a single computation-efficient GPU, while offloading cold neurons to NDP-DIMMs, which offer large memory size but limited computation capabilities. Meanwhile, the dynamic nature of activation sparsity needs a real-time partition of hot/cold neurons and adaptive remapping of cold neurons across multiple NDP-DIMM modules. Therefore, we introduce a lightweight predictor optimizing real-time neuron partition and adjustment between GPU and NDP-DIMMs. We also utilize a window-based online scheduling mechanism to maintain load balance among NDP-DIMM modules. Hermes facilitates the deployment of LLaMA2-70B on consumer-grade hardware at 13.75 tokens/s and realizes an average 75.24$\times$ speedup over the state-of-the-art offloading-based inference system. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16963v1-abstract-full').style.display = 'none'; document.getElementById('2502.16963v1-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> <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, 17 figures, accepted by HPCA 2025</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> C.1.3 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.16140">arXiv:2502.16140</a> <span> [<a href="https://arxiv.org/pdf/2502.16140">pdf</a>, <a href="https://arxiv.org/format/2502.16140">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> </div> </div> <p class="title is-5 mathjax"> Semantic Gaussian Mixture Variational Autoencoder for Sequential Recommendation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Beibei Li</a>, <a href="/search/cs?searchtype=author&query=Xiang%2C+T">Tao Xiang</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+B">Beihong Jin</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+Y">Yiyuan Zheng</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+R">Rui Zhao</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.16140v1-abstract-short" style="display: inline;"> Variational AutoEncoder (VAE) for Sequential Recommendation (SR), which learns a continuous distribution for each user-item interaction sequence rather than a determinate embedding, is robust against data deficiency and achieves significant performance. However, existing VAE-based SR models assume a unimodal Gaussian distribution as the prior distribution of sequence representations, leading to re… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16140v1-abstract-full').style.display = 'inline'; document.getElementById('2502.16140v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.16140v1-abstract-full" style="display: none;"> Variational AutoEncoder (VAE) for Sequential Recommendation (SR), which learns a continuous distribution for each user-item interaction sequence rather than a determinate embedding, is robust against data deficiency and achieves significant performance. However, existing VAE-based SR models assume a unimodal Gaussian distribution as the prior distribution of sequence representations, leading to restricted capability to capture complex user interests and limiting recommendation performance when users have more than one interest. Due to that it is common for users to have multiple disparate interests, we argue that it is more reasonable to establish a multimodal prior distribution in SR scenarios instead of a unimodal one. Therefore, in this paper, we propose a novel VAE-based SR model named SIGMA. SIGMA assumes that the prior of sequence representation conforms to a Gaussian mixture distribution, where each component of the distribution semantically corresponds to one of multiple interests. For multi-interest elicitation, SIGMA includes a probabilistic multi-interest extraction module that learns a unimodal Gaussian distribution for each interest according to implicit item hyper-categories. Additionally, to incorporate the multimodal interests into sequence representation learning, SIGMA constructs a multi-interest-aware ELBO, which is compatible with the Gaussian mixture prior. Extensive experiments on public datasets demonstrate the effectiveness of SIGMA. The code is available at https://github.com/libeibei95/SIGMA. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.16140v1-abstract-full').style.display = 'none'; document.getElementById('2502.16140v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 22 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">Accepted by DASFAA 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/2502.15902">arXiv:2502.15902</a> <span> [<a href="https://arxiv.org/pdf/2502.15902">pdf</a>, <a href="https://arxiv.org/format/2502.15902">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"> IPAD: Inverse Prompt for AI Detection -- A Robust and Explainable LLM-Generated Text Detector </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+Z">Zheng Chen</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+Y">Yushi Feng</a>, <a href="/search/cs?searchtype=author&query=He%2C+C">Changyang He</a>, <a href="/search/cs?searchtype=author&query=Deng%2C+Y">Yue Deng</a>, <a href="/search/cs?searchtype=author&query=Pu%2C+H">Hongxi Pu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo 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="2502.15902v1-abstract-short" style="display: inline;"> Large Language Models (LLMs) have attained human-level fluency in text generation, which complicates the distinguishing between human-written and LLM-generated texts. This increases the risk of misuse and highlights the need for reliable detectors. Yet, existing detectors exhibit poor robustness on out-of-distribution (OOD) data and attacked data, which is critical for real-world scenarios. Also,… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15902v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15902v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15902v1-abstract-full" style="display: none;"> Large Language Models (LLMs) have attained human-level fluency in text generation, which complicates the distinguishing between human-written and LLM-generated texts. This increases the risk of misuse and highlights the need for reliable detectors. Yet, existing detectors exhibit poor robustness on out-of-distribution (OOD) data and attacked data, which is critical for real-world scenarios. Also, they struggle to provide explainable evidence to support their decisions, thus undermining the reliability. In light of these challenges, we propose IPAD (Inverse Prompt for AI Detection), a novel framework consisting of a Prompt Inverter that identifies predicted prompts that could have generated the input text, and a Distinguisher that examines how well the input texts align with the predicted prompts. We develop and examine two versions of Distinguishers. Empirical evaluations demonstrate that both Distinguishers perform significantly better than the baseline methods, with version2 outperforming baselines by 9.73% on in-distribution data (F1-score) and 12.65% on OOD data (AUROC). Furthermore, a user study is conducted to illustrate that IPAD enhances the AI detection trustworthiness by allowing users to directly examine the decision-making evidence, which provides interpretable support for its state-of-the-art detection results. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15902v1-abstract-full').style.display = 'none'; document.getElementById('2502.15902v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.15803">arXiv:2502.15803</a> <span> [<a href="https://arxiv.org/pdf/2502.15803">pdf</a>, <a href="https://arxiv.org/format/2502.15803">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"> Megrez-Omni Technical Report </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Boxun Li</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yadong Li</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Z">Zhiyuan Li</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+C">Congyi Liu</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+W">Weilin Liu</a>, <a href="/search/cs?searchtype=author&query=Niu%2C+G">Guowei Niu</a>, <a href="/search/cs?searchtype=author&query=Tan%2C+Z">Zheyue Tan</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+H">Haiyang Xu</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+Z">Zhuyu Yao</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+T">Tao Yuan</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+D">Dong Zhou</a>, <a href="/search/cs?searchtype=author&query=Zhuang%2C+Y">Yueqing Zhuang</a>, <a href="/search/cs?searchtype=author&query=Yan%2C+S">Shengen Yan</a>, <a href="/search/cs?searchtype=author&query=Dai%2C+G">Guohao Dai</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yu 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="2502.15803v1-abstract-short" style="display: inline;"> In this work, we present the Megrez models, comprising a language model (Megrez-3B-Instruct) and a multimodal model (Megrez-3B-Omni). These models are designed to deliver fast inference, compactness, and robust edge-side intelligence through a software-hardware co-design approach. Megrez-3B-Instruct offers several advantages, including high accuracy, high speed, ease of use, and a wide range of ap… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15803v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15803v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15803v1-abstract-full" style="display: none;"> In this work, we present the Megrez models, comprising a language model (Megrez-3B-Instruct) and a multimodal model (Megrez-3B-Omni). These models are designed to deliver fast inference, compactness, and robust edge-side intelligence through a software-hardware co-design approach. Megrez-3B-Instruct offers several advantages, including high accuracy, high speed, ease of use, and a wide range of applications. Building on Megrez-3B-Instruct, Megrez-3B-Omni is an on-device multimodal understanding LLM that supports image, text, and audio analysis. It achieves state-of-the-art accuracy across all three modalities and demonstrates strong versatility and robustness, setting a new benchmark for multimodal AI models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15803v1-abstract-full').style.display = 'none'; document.getElementById('2502.15803v1-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 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/2502.15466">arXiv:2502.15466</a> <span> [<a href="https://arxiv.org/pdf/2502.15466">pdf</a>, <a href="https://arxiv.org/format/2502.15466">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"> Mitigating Data Scarcity in Time Series Analysis: A Foundation Model with Series-Symbol Data Generation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+W">Wenxuan Wang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+K">Kai Wu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y+B">Yujian Betterest Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+D">Dan Wang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+X">Xiaoyu Zhang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jing Liu</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.15466v1-abstract-short" style="display: inline;"> Foundation models for time series analysis (TSA) have attracted significant attention. However, challenges such as data scarcity and data imbalance continue to hinder their development. To address this, we consider modeling complex systems through symbolic expressions that serve as semantic descriptors of time series. Building on this concept, we introduce a series-symbol (S2) dual-modulity data g… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15466v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15466v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15466v1-abstract-full" style="display: none;"> Foundation models for time series analysis (TSA) have attracted significant attention. However, challenges such as data scarcity and data imbalance continue to hinder their development. To address this, we consider modeling complex systems through symbolic expressions that serve as semantic descriptors of time series. Building on this concept, we introduce a series-symbol (S2) dual-modulity data generation mechanism, enabling the unrestricted creation of high-quality time series data paired with corresponding symbolic representations. Leveraging the S2 dataset, we develop SymTime, a pre-trained foundation model for TSA. SymTime demonstrates competitive performance across five major TSA tasks when fine-tuned with downstream task, rivaling foundation models pre-trained on real-world datasets. This approach underscores the potential of dual-modality data generation and pretraining mechanisms in overcoming data scarcity and enhancing task performance. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15466v1-abstract-full').style.display = 'none'; document.getElementById('2502.15466v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.15292">arXiv:2502.15292</a> <span> [<a href="https://arxiv.org/pdf/2502.15292">pdf</a>, <a href="https://arxiv.org/format/2502.15292">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Software Engineering">cs.SE</span> </div> </div> <p class="title is-5 mathjax"> Bridging Bug Localization and Issue Fixing: A Hierarchical Localization Framework Leveraging Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chang%2C+J">Jianming Chang</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+X">Xin Zhou</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+L">Lulu Wang</a>, <a href="/search/cs?searchtype=author&query=Lo%2C+D">David Lo</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bixin 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="2502.15292v1-abstract-short" style="display: inline;"> Automated issue fixing is a critical task in software debugging and has recently garnered significant attention from academia and industry. However, existing fixing techniques predominantly focus on the repair phase, often overlooking the importance of improving the preceding bug localization phase. As a foundational step in issue fixing, bug localization plays a pivotal role in determining the ov… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15292v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15292v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15292v1-abstract-full" style="display: none;"> Automated issue fixing is a critical task in software debugging and has recently garnered significant attention from academia and industry. However, existing fixing techniques predominantly focus on the repair phase, often overlooking the importance of improving the preceding bug localization phase. As a foundational step in issue fixing, bug localization plays a pivotal role in determining the overall effectiveness of the entire process. To enhance the precision of issue fixing by accurately identifying bug locations in large-scale projects, this paper presents BugCerberus, the first hierarchical bug localization framework powered by three customized large language models. First, BugCerberus analyzes intermediate representations of bug-related programs at file, function, and statement levels and extracts bug-related contextual information from the representations. Second, BugCerberus designs three customized LLMs at each level using bug reports and contexts to learn the patterns of bugs. Finally, BugCerberus hierarchically searches for bug-related code elements through well-tuned models to localize bugs at three levels. With BugCerberus, we further investigate the impact of bug localization on the issue fixing. We evaluate BugCerberus on the widely-used benchmark SWE-bench-lite. The experimental results demonstrate that BugCerberus outperforms all baselines. Specifically, at the fine-grained statement level, BugCerberus surpasses the state-of-the-art in Top-N (N=1, 3, 5, 10) by 16.5%, 5.4%, 10.2%, and 23.1%, respectively. Moreover, in the issue fixing experiments, BugCerberus improves the fix rate of the existing issue fixing approach Agentless by 17.4% compared to the best baseline, highlighting the significant impact of enhanced bug localization on automated issue fixing. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15292v1-abstract-full').style.display = 'none'; document.getElementById('2502.15292v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.14931">arXiv:2502.14931</a> <span> [<a href="https://arxiv.org/pdf/2502.14931">pdf</a>, <a href="https://arxiv.org/format/2502.14931">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"> Hier-SLAM++: Neuro-Symbolic Semantic SLAM with a Hierarchically Categorical Gaussian Splatting </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Boying Li</a>, <a href="/search/cs?searchtype=author&query=Hao%2C+V+C">Vuong Chi Hao</a>, <a href="/search/cs?searchtype=author&query=Stuckey%2C+P+J">Peter J. Stuckey</a>, <a href="/search/cs?searchtype=author&query=Reid%2C+I">Ian Reid</a>, <a href="/search/cs?searchtype=author&query=Rezatofighi%2C+H">Hamid Rezatofighi</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="2502.14931v1-abstract-short" style="display: inline;"> We propose Hier-SLAM++, a comprehensive Neuro-Symbolic semantic 3D Gaussian Splatting SLAM method with both RGB-D and monocular input featuring an advanced hierarchical categorical representation, which enables accurate pose estimation as well as global 3D semantic mapping. The parameter usage in semantic SLAM systems increases significantly with the growing complexity of the environment, making s… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14931v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14931v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14931v1-abstract-full" style="display: none;"> We propose Hier-SLAM++, a comprehensive Neuro-Symbolic semantic 3D Gaussian Splatting SLAM method with both RGB-D and monocular input featuring an advanced hierarchical categorical representation, which enables accurate pose estimation as well as global 3D semantic mapping. The parameter usage in semantic SLAM systems increases significantly with the growing complexity of the environment, making scene understanding particularly challenging and costly. To address this problem, we introduce a novel and general hierarchical representation that encodes both semantic and geometric information in a compact form into 3D Gaussian Splatting, leveraging the capabilities of large language models (LLMs) as well as the 3D generative model. By utilizing the proposed hierarchical tree structure, semantic information is symbolically represented and learned in an end-to-end manner. We further introduce a novel semantic loss designed to optimize hierarchical semantic information through both inter-level and cross-level optimization. Additionally, we propose an improved SLAM system to support both RGB-D and monocular inputs using a feed-forward model. To the best of our knowledge, this is the first semantic monocular Gaussian Splatting SLAM system, significantly reducing sensor requirements for 3D semantic understanding and broadening the applicability of semantic Gaussian SLAM system. We conduct experiments on both synthetic and real-world datasets, demonstrating superior or on-par performance with state-of-the-art NeRF-based and Gaussian-based SLAM systems, while significantly reducing storage and training time requirements. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14931v1-abstract-full').style.display = 'none'; document.getElementById('2502.14931v1-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> <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. Under review</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.14922">arXiv:2502.14922</a> <span> [<a href="https://arxiv.org/pdf/2502.14922">pdf</a>, <a href="https://arxiv.org/format/2502.14922">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"> SIFT: Grounding LLM Reasoning in Contexts via Stickers </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zeng%2C+Z">Zihao Zeng</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+X">Xuyao Huang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Boxiu Li</a>, <a href="/search/cs?searchtype=author&query=Deng%2C+Z">Zhijie Deng</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="2502.14922v1-abstract-short" style="display: inline;"> This paper identifies the misinterpretation of the context can be a significant issue during the reasoning process of large language models, spanning from smaller models like Llama3.2-3B-Instruct to cutting-edge ones like DeepSeek-R1. For example, in the phrase "10 dollars per kilo," LLMs might not recognize that "per" means "for each," leading to calculation errors. We introduce a novel, post-tra… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14922v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14922v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14922v1-abstract-full" style="display: none;"> This paper identifies the misinterpretation of the context can be a significant issue during the reasoning process of large language models, spanning from smaller models like Llama3.2-3B-Instruct to cutting-edge ones like DeepSeek-R1. For example, in the phrase "10 dollars per kilo," LLMs might not recognize that "per" means "for each," leading to calculation errors. We introduce a novel, post-training approach called **Stick to the Facts (SIFT)** to tackle this. SIFT leverages increasing inference-time compute to ground LLM reasoning in contexts. At the core of SIFT lies the *Sticker*, which is generated by the model itself to explicitly emphasize the key information within the context. Given the curated Sticker, SIFT generates two predictions -- one from the original query and one from the query augmented with the Sticker. If they differ, the Sticker is sequentially refined via *forward* optimization (to better align the extracted facts with the query) and *inverse* generation (to conform with the model's inherent tendencies) for more faithful reasoning outcomes. Studies across diverse models (from 3B to 100B+) and benchmarks (e.g., GSM8K, MATH-500) reveal consistent performance improvements. Notably, SIFT improves the pass@1 accuracy of DeepSeek-R1 on AIME2024 from 78.33% to **85.67**%, establishing a new state-of-the-art in the open-source community. The code is available at https://github.com/zhijie-group/SIFT. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14922v1-abstract-full').style.display = 'none'; document.getElementById('2502.14922v1-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 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/2502.14340">arXiv:2502.14340</a> <span> [<a href="https://arxiv.org/pdf/2502.14340">pdf</a>, <a href="https://arxiv.org/format/2502.14340">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"> Earlier Tokens Contribute More: Learning Direct Preference Optimization From Temporal Decay Perspective </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shao%2C+R">Ruichen Shao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bei Li</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+G">Gangao Liu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yang Chen</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+X">Xiang Zhou</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Jingang Wang</a>, <a href="/search/cs?searchtype=author&query=Cai%2C+X">Xunliang Cai</a>, <a href="/search/cs?searchtype=author&query=Li%2C+P">Peng 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="2502.14340v1-abstract-short" style="display: inline;"> Direct Preference Optimization (DPO) has gained attention as an efficient alternative to reinforcement learning from human feedback (RLHF) for aligning large language models (LLMs) with human preferences. Despite its advantages, DPO suffers from a length bias, generating responses longer than those from the reference model. Existing solutions like SimPO and SamPO address this issue but uniformly t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14340v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14340v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14340v1-abstract-full" style="display: none;"> Direct Preference Optimization (DPO) has gained attention as an efficient alternative to reinforcement learning from human feedback (RLHF) for aligning large language models (LLMs) with human preferences. Despite its advantages, DPO suffers from a length bias, generating responses longer than those from the reference model. Existing solutions like SimPO and SamPO address this issue but uniformly treat the contribution of rewards across sequences, overlooking temporal dynamics. To this end, we propose an enhanced preference optimization method that incorporates a temporal decay factor controlled by a gamma parameter. This dynamic weighting mechanism adjusts the influence of each reward based on its position in the sequence, prioritizing earlier tokens that are more critical for alignment. By adaptively focusing on more relevant feedback, our approach mitigates overfitting to less pertinent data and remains responsive to evolving human preferences. Experimental results on several benchmarks show that our approach consistently outperforms vanilla DPO by 5.9-8.8 points on AlpacaEval 2 and 3.3-9.7 points on Arena-Hard across different model architectures and sizes. Furthermore, additional experiments on mathematical and reasoning benchmarks (MMLU, GSM8K, and MATH) confirm that our method enhances performance without compromising general capabilities. Our codebase would be available at \url{https://github.com/LotuSrc/D2PO}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14340v1-abstract-full').style.display = 'none'; document.getElementById('2502.14340v1-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> <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 ICLR 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/2502.14296">arXiv:2502.14296</a> <span> [<a href="https://arxiv.org/pdf/2502.14296">pdf</a>, <a href="https://arxiv.org/format/2502.14296">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> </div> </div> <p class="title is-5 mathjax"> On the Trustworthiness of Generative Foundation Models: Guideline, Assessment, and Perspective </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yue Huang</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+C">Chujie Gao</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+S">Siyuan Wu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Haoran Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xiangqi Wang</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Y">Yujun Zhou</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yanbo Wang</a>, <a href="/search/cs?searchtype=author&query=Ye%2C+J">Jiayi Ye</a>, <a href="/search/cs?searchtype=author&query=Shi%2C+J">Jiawen Shi</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Q">Qihui Zhang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yuan Li</a>, <a href="/search/cs?searchtype=author&query=Bao%2C+H">Han Bao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Z">Zhaoyi Liu</a>, <a href="/search/cs?searchtype=author&query=Guan%2C+T">Tianrui Guan</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+D">Dongping Chen</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+R">Ruoxi Chen</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+K">Kehan Guo</a>, <a href="/search/cs?searchtype=author&query=Zou%2C+A">Andy Zou</a>, <a href="/search/cs?searchtype=author&query=Kuen-Yew%2C+B+H">Bryan Hooi Kuen-Yew</a>, <a href="/search/cs?searchtype=author&query=Xiong%2C+C">Caiming Xiong</a>, <a href="/search/cs?searchtype=author&query=Stengel-Eskin%2C+E">Elias Stengel-Eskin</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Hongyang Zhang</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+H">Hongzhi Yin</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Huan Zhang</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+H">Huaxiu Yao</a> , et al. (41 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="2502.14296v1-abstract-short" style="display: inline;"> Generative Foundation Models (GenFMs) have emerged as transformative tools. However, their widespread adoption raises critical concerns regarding trustworthiness across dimensions. This paper presents a comprehensive framework to address these challenges through three key contributions. First, we systematically review global AI governance laws and policies from governments and regulatory bodies, a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14296v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14296v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14296v1-abstract-full" style="display: none;"> Generative Foundation Models (GenFMs) have emerged as transformative tools. However, their widespread adoption raises critical concerns regarding trustworthiness across dimensions. This paper presents a comprehensive framework to address these challenges through three key contributions. First, we systematically review global AI governance laws and policies from governments and regulatory bodies, as well as industry practices and standards. Based on this analysis, we propose a set of guiding principles for GenFMs, developed through extensive multidisciplinary collaboration that integrates technical, ethical, legal, and societal perspectives. Second, we introduce TrustGen, the first dynamic benchmarking platform designed to evaluate trustworthiness across multiple dimensions and model types, including text-to-image, large language, and vision-language models. TrustGen leverages modular components--metadata curation, test case generation, and contextual variation--to enable adaptive and iterative assessments, overcoming the limitations of static evaluation methods. Using TrustGen, we reveal significant progress in trustworthiness while identifying persistent challenges. Finally, we provide an in-depth discussion of the challenges and future directions for trustworthy GenFMs, which reveals the complex, evolving nature of trustworthiness, highlighting the nuanced trade-offs between utility and trustworthiness, and consideration for various downstream applications, identifying persistent challenges and providing a strategic roadmap for future research. This work establishes a holistic framework for advancing trustworthiness in GenAI, paving the way for safer and more responsible integration of GenFMs into critical applications. To facilitate advancement in the community, we release the toolkit for dynamic evaluation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14296v1-abstract-full').style.display = 'none'; document.getElementById('2502.14296v1-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/2502.14129">arXiv:2502.14129</a> <span> [<a href="https://arxiv.org/pdf/2502.14129">pdf</a>, <a href="https://arxiv.org/format/2502.14129">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"> GlossGau: Efficient Inverse Rendering for Glossy Surface with Anisotropic Spherical Gaussian </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Du%2C+B">Bang Du</a>, <a href="/search/cs?searchtype=author&query=Li%2C+R+B">Runfa Blark Li</a>, <a href="/search/cs?searchtype=author&query=Du%2C+C">Chen Du</a>, <a href="/search/cs?searchtype=author&query=Nguyen%2C+T">Truong Nguyen</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="2502.14129v1-abstract-short" style="display: inline;"> The reconstruction of 3D objects from calibrated photographs represents a fundamental yet intricate challenge in the domains of computer graphics and vision. Although neural reconstruction approaches based on Neural Radiance Fields (NeRF) have shown remarkable capabilities, their processing costs remain substantial. Recently, the advent of 3D Gaussian Splatting (3D-GS) largely improves the trainin… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14129v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14129v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14129v1-abstract-full" style="display: none;"> The reconstruction of 3D objects from calibrated photographs represents a fundamental yet intricate challenge in the domains of computer graphics and vision. Although neural reconstruction approaches based on Neural Radiance Fields (NeRF) have shown remarkable capabilities, their processing costs remain substantial. Recently, the advent of 3D Gaussian Splatting (3D-GS) largely improves the training efficiency and facilitates to generate realistic rendering in real-time. However, due to the limited ability of Spherical Harmonics (SH) to represent high-frequency information, 3D-GS falls short in reconstructing glossy objects. Researchers have turned to enhance the specular expressiveness of 3D-GS through inverse rendering. Yet these methods often struggle to maintain the training and rendering efficiency, undermining the benefits of Gaussian Splatting techniques. In this paper, we introduce GlossGau, an efficient inverse rendering framework that reconstructs scenes with glossy surfaces while maintaining training and rendering speeds comparable to vanilla 3D-GS. Specifically, we explicitly model the surface normals, Bidirectional Reflectance Distribution Function (BRDF) parameters, as well as incident lights and use Anisotropic Spherical Gaussian (ASG) to approximate the per-Gaussian Normal Distribution Function under the microfacet model. We utilize 2D Gaussian Splatting (2D-GS) as foundational primitives and apply regularization to significantly alleviate the normal estimation challenge encountered in related works. Experiments demonstrate that GlossGau achieves competitive or superior reconstruction on datasets with glossy surfaces. Compared with previous GS-based works that address the specular surface, our optimization time is considerably less. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14129v1-abstract-full').style.display = 'none'; document.getElementById('2502.14129v1-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 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/2502.13671">arXiv:2502.13671</a> <span> [<a href="https://arxiv.org/pdf/2502.13671">pdf</a>, <a href="https://arxiv.org/format/2502.13671">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> </div> </div> <p class="title is-5 mathjax"> On the Subsidy of Envy-Free Orientations in Graphs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+A">Ankang Sun</a>, <a href="/search/cs?searchtype=author&query=Suzuki%2C+M">Mashbat Suzuki</a>, <a href="/search/cs?searchtype=author&query=Xing%2C+S">Shiji Xing</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="2502.13671v1-abstract-short" style="display: inline;"> We study a fair division problem in (multi)graphs where $n$ agents (vertices) are pairwise connected by items (edges), and each agent is only interested in its incident items. We consider how to allocate items to incident agents in an envy-free manner, i.e., envy-free orientations, while minimizing the overall payment, i.e., subsidy. We first prove that computing an envy-free orientation with the… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13671v1-abstract-full').style.display = 'inline'; document.getElementById('2502.13671v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.13671v1-abstract-full" style="display: none;"> We study a fair division problem in (multi)graphs where $n$ agents (vertices) are pairwise connected by items (edges), and each agent is only interested in its incident items. We consider how to allocate items to incident agents in an envy-free manner, i.e., envy-free orientations, while minimizing the overall payment, i.e., subsidy. We first prove that computing an envy-free orientation with the minimum subsidy is NP-hard, even when the graph is simple and the agents have bi-valued additive valuations. We then bound the worst-case subsidy. We prove that for any multigraph (i.e., allowing parallel edges) and monotone valuations where the marginal value of each good is at most \$1 for each agent, \$1 each (a total subsidy of $n-1$, where $n$ is the number of agents) is sufficient. This is one of the few cases where linear subsidy $螛(n)$ is known to be necessary and sufficient to guarantee envy-freeness when agents have monotone valuations. When the valuations are additive (while the graph may contain parallel edges) and when the graph is simple (while the valuations may be monotone), we improve the bound to $n/2$ and $n-2$, respectively. Moreover, these two bounds are tight. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13671v1-abstract-full').style.display = 'none'; document.getElementById('2502.13671v1-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 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/2502.13143">arXiv:2502.13143</a> <span> [<a href="https://arxiv.org/pdf/2502.13143">pdf</a>, <a href="https://arxiv.org/format/2502.13143">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> <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"> SoFar: Language-Grounded Orientation Bridges Spatial Reasoning and Object Manipulation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Qi%2C+Z">Zekun Qi</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wenyao Zhang</a>, <a href="/search/cs?searchtype=author&query=Ding%2C+Y">Yufei Ding</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+R">Runpei Dong</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+X">Xinqiang Yu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+J">Jingwen Li</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+L">Lingyun Xu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Baoyu Li</a>, <a href="/search/cs?searchtype=author&query=He%2C+X">Xialin He</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+G">Guofan Fan</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jiazhao Zhang</a>, <a href="/search/cs?searchtype=author&query=He%2C+J">Jiawei He</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+J">Jiayuan Gu</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+X">Xin Jin</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+K">Kaisheng Ma</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhizheng Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">He Wang</a>, <a href="/search/cs?searchtype=author&query=Yi%2C+L">Li Yi</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="2502.13143v1-abstract-short" style="display: inline;"> Spatial intelligence is a critical component of embodied AI, promoting robots to understand and interact with their environments. While recent advances have enhanced the ability of VLMs to perceive object locations and positional relationships, they still lack the capability to precisely understand object orientations-a key requirement for tasks involving fine-grained manipulations. Addressing thi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13143v1-abstract-full').style.display = 'inline'; document.getElementById('2502.13143v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.13143v1-abstract-full" style="display: none;"> Spatial intelligence is a critical component of embodied AI, promoting robots to understand and interact with their environments. While recent advances have enhanced the ability of VLMs to perceive object locations and positional relationships, they still lack the capability to precisely understand object orientations-a key requirement for tasks involving fine-grained manipulations. Addressing this limitation not only requires geometric reasoning but also an expressive and intuitive way to represent orientation. In this context, we propose that natural language offers a more flexible representation space than canonical frames, making it particularly suitable for instruction-following robotic systems. In this paper, we introduce the concept of semantic orientation, which defines object orientations using natural language in a reference-frame-free manner (e.g., the ''plug-in'' direction of a USB or the ''handle'' direction of a knife). To support this, we construct OrienText300K, a large-scale dataset of 3D models annotated with semantic orientations that link geometric understanding to functional semantics. By integrating semantic orientation into a VLM system, we enable robots to generate manipulation actions with both positional and orientational constraints. Extensive experiments in simulation and real world demonstrate that our approach significantly enhances robotic manipulation capabilities, e.g., 48.7% accuracy on Open6DOR and 74.9% accuracy on SIMPLER. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13143v1-abstract-full').style.display = 'none'; document.getElementById('2502.13143v1-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 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">Project page: https://qizekun.github.io/sofar/</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.12608">arXiv:2502.12608</a> <span> [<a href="https://arxiv.org/pdf/2502.12608">pdf</a>, <a href="https://arxiv.org/format/2502.12608">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"> Unveiling Mode Connectivity in Graph Neural Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Bingheng Li</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Z">Zhikai Chen</a>, <a href="/search/cs?searchtype=author&query=Han%2C+H">Haoyu Han</a>, <a href="/search/cs?searchtype=author&query=Zeng%2C+S">Shenglai Zeng</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jingzhe Liu</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+J">Jiliang Tang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.12608v1-abstract-short" style="display: inline;"> A fundamental challenge in understanding graph neural networks (GNNs) lies in characterizing their optimization dynamics and loss landscape geometry, critical for improving interpretability and robustness. While mode connectivity, a lens for analyzing geometric properties of loss landscapes has proven insightful for other deep learning architectures, its implications for GNNs remain unexplored. Th… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12608v1-abstract-full').style.display = 'inline'; document.getElementById('2502.12608v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.12608v1-abstract-full" style="display: none;"> A fundamental challenge in understanding graph neural networks (GNNs) lies in characterizing their optimization dynamics and loss landscape geometry, critical for improving interpretability and robustness. While mode connectivity, a lens for analyzing geometric properties of loss landscapes has proven insightful for other deep learning architectures, its implications for GNNs remain unexplored. This work presents the first investigation of mode connectivity in GNNs. We uncover that GNNs exhibit distinct non-linear mode connectivity, diverging from patterns observed in fully-connected networks or CNNs. Crucially, we demonstrate that graph structure, rather than model architecture, dominates this behavior, with graph properties like homophily correlating with mode connectivity patterns. We further establish a link between mode connectivity and generalization, proposing a generalization bound based on loss barriers and revealing its utility as a diagnostic tool. Our findings further bridge theoretical insights with practical implications: they rationalize domain alignment strategies in graph learning and provide a foundation for refining GNN training paradigms. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12608v1-abstract-full').style.display = 'none'; document.getElementById('2502.12608v1-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 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/2502.12221">arXiv:2502.12221</a> <span> [<a href="https://arxiv.org/pdf/2502.12221">pdf</a>, <a href="https://arxiv.org/format/2502.12221">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Software Engineering">cs.SE</span> </div> </div> <p class="title is-5 mathjax"> ReF Decompile: Relabeling and Function Call Enhanced Decompile </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Feng%2C+Y">Yunlong Feng</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bohan Li</a>, <a href="/search/cs?searchtype=author&query=Shi%2C+X">Xiaoming Shi</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+Q">Qingfu Zhu</a>, <a href="/search/cs?searchtype=author&query=Che%2C+W">Wanxiang Che</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="2502.12221v1-abstract-short" style="display: inline;"> The goal of decompilation is to convert compiled low-level code (e.g., assembly code) back into high-level programming languages, enabling analysis in scenarios where source code is unavailable. This task supports various reverse engineering applications, such as vulnerability identification, malware analysis, and legacy software migration. The end-to-end decompile method based on large langauge m… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12221v1-abstract-full').style.display = 'inline'; document.getElementById('2502.12221v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.12221v1-abstract-full" style="display: none;"> The goal of decompilation is to convert compiled low-level code (e.g., assembly code) back into high-level programming languages, enabling analysis in scenarios where source code is unavailable. This task supports various reverse engineering applications, such as vulnerability identification, malware analysis, and legacy software migration. The end-to-end decompile method based on large langauge models (LLMs) reduces reliance on additional tools and minimizes manual intervention due to its inherent properties. However, previous end-to-end methods often lose critical information necessary for reconstructing control flow structures and variables when processing binary files, making it challenging to accurately recover the program's logic. To address these issues, we propose the \textbf{ReF Decompile} method, which incorporates the following innovations: (1) The Relabelling strategy replaces jump target addresses with labels, preserving control flow clarity. (2) The Function Call strategy infers variable types and retrieves missing variable information from binary files. Experimental results on the Humaneval-Decompile Benchmark demonstrate that ReF Decompile surpasses comparable baselines and achieves state-of-the-art (SOTA) performance of $61.43\%$. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12221v1-abstract-full').style.display = 'none'; document.getElementById('2502.12221v1-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/2502.12217">arXiv:2502.12217</a> <span> [<a href="https://arxiv.org/pdf/2502.12217">pdf</a>, <a href="https://arxiv.org/format/2502.12217">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"> Optimal Brain Iterative Merging: Mitigating Interference in LLM Merging </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhixiang Wang</a>, <a href="/search/cs?searchtype=author&query=Mao%2C+Z">Zhenyu Mao</a>, <a href="/search/cs?searchtype=author&query=Qiao%2C+Y">Yixuan Qiao</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+Y">Yunfang Wu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Biye 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="2502.12217v1-abstract-short" style="display: inline;"> Large Language Models (LLMs) have demonstrated impressive capabilities, but their high computational costs pose challenges for customization. Model merging offers a cost-effective alternative, yet existing methods suffer from interference among parameters, leading to performance degradation. In this work, we propose Optimal Brain Iterative Merging (OBIM), a novel method designed to mitigate both i… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12217v1-abstract-full').style.display = 'inline'; document.getElementById('2502.12217v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.12217v1-abstract-full" style="display: none;"> Large Language Models (LLMs) have demonstrated impressive capabilities, but their high computational costs pose challenges for customization. Model merging offers a cost-effective alternative, yet existing methods suffer from interference among parameters, leading to performance degradation. In this work, we propose Optimal Brain Iterative Merging (OBIM), a novel method designed to mitigate both intra-model and inter-model interference. OBIM consists of two key components: (1) A saliency measurement mechanism that evaluates parameter importance based on loss changes induced by individual weight alterations, reducing intra-model interference by preserving only high-saliency parameters. (2) A mutually exclusive iterative merging framework, which incrementally integrates models using a binary mask to avoid direct parameter averaging, thereby mitigating inter-model interference. We validate OBIM through experiments on both Supervised Fine-Tuned (SFT) models and post-pretrained checkpoints. The results show that OBIM significantly outperforms existing merging techniques. Overall, OBIM provides an effective and practical solution for enhancing LLM merging. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12217v1-abstract-full').style.display = 'none'; document.getElementById('2502.12217v1-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/2502.12025">arXiv:2502.12025</a> <span> [<a href="https://arxiv.org/pdf/2502.12025">pdf</a>, <a href="https://arxiv.org/format/2502.12025">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jiang%2C+F">Fengqing Jiang</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+Z">Zhangchen Xu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yuetai Li</a>, <a href="/search/cs?searchtype=author&query=Niu%2C+L">Luyao Niu</a>, <a href="/search/cs?searchtype=author&query=Xiang%2C+Z">Zhen Xiang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+B+Y">Bill Yuchen Lin</a>, <a href="/search/cs?searchtype=author&query=Poovendran%2C+R">Radha Poovendran</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="2502.12025v1-abstract-short" style="display: inline;"> Emerging large reasoning models (LRMs), such as DeepSeek-R1 models, leverage long chain-of-thought (CoT) reasoning to generate structured intermediate steps, enhancing their reasoning capabilities. However, long CoT does not inherently guarantee safe outputs, potentially leading to harmful consequences such as the introduction of security vulnerabilities in code or the spread of misinformation. Cu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12025v1-abstract-full').style.display = 'inline'; document.getElementById('2502.12025v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.12025v1-abstract-full" style="display: none;"> Emerging large reasoning models (LRMs), such as DeepSeek-R1 models, leverage long chain-of-thought (CoT) reasoning to generate structured intermediate steps, enhancing their reasoning capabilities. However, long CoT does not inherently guarantee safe outputs, potentially leading to harmful consequences such as the introduction of security vulnerabilities in code or the spread of misinformation. Current research on large language model (LLM) safety usually focuses on short-answer responses, overlooking the long CoT style outputs of LRMs. To bridge this gap, we conduct a systematic study of LRM safety. First, we investigate safety evaluators calibrated against human annotations. Using our newly developed metrics, we thoroughly assess the safety of 12 state-of-the-art LRMs on StrongReject and WildJailbreak datasets. Our results show that LRMs are not safe compared to their reasoning advance. Further, we perform a fine-grained analysis of the reasoning trace and final answer. We find that three decoding strategies-ZeroThink, LessThink, and MoreThink-can improve model safety without additional training. However, these strategies either use constrained reasoning traces or incur high inference costs. To better strengthen LRM safety, we introduce SafeChain, the first-of-its-kind safety training dataset in CoT style. We fine-tune two LRMs with SafeChain, showing that it not only enhances model safety but also preserves performance across 6 reasoning benchmarks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12025v1-abstract-full').style.display = 'none'; document.getElementById('2502.12025v1-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/2502.11946">arXiv:2502.11946</a> <span> [<a href="https://arxiv.org/pdf/2502.11946">pdf</a>, <a href="https://arxiv.org/format/2502.11946">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="Human-Computer Interaction">cs.HC</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"> Step-Audio: Unified Understanding and Generation in Intelligent Speech Interaction </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Huang%2C+A">Ailin Huang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+B">Boyong Wu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+B">Bruce Wang</a>, <a href="/search/cs?searchtype=author&query=Yan%2C+C">Chao Yan</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+C">Chen Hu</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+C">Chengli Feng</a>, <a href="/search/cs?searchtype=author&query=Tian%2C+F">Fei Tian</a>, <a href="/search/cs?searchtype=author&query=Shen%2C+F">Feiyu Shen</a>, <a href="/search/cs?searchtype=author&query=Li%2C+J">Jingbei Li</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+M">Mingrui Chen</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+P">Peng Liu</a>, <a href="/search/cs?searchtype=author&query=Miao%2C+R">Ruihang Miao</a>, <a href="/search/cs?searchtype=author&query=You%2C+W">Wang You</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xi Chen</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+X">Xuerui Yang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yechang Huang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yuxiang Zhang</a>, <a href="/search/cs?searchtype=author&query=Gong%2C+Z">Zheng Gong</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zixin Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+H">Hongyu Zhou</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+J">Jianjian Sun</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Brian Li</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+C">Chengting Feng</a>, <a href="/search/cs?searchtype=author&query=Wan%2C+C">Changyi Wan</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+H">Hanpeng Hu</a> , et al. (120 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="2502.11946v2-abstract-short" style="display: inline;"> Real-time speech interaction, serving as a fundamental interface for human-machine collaboration, holds immense potential. However, current open-source models face limitations such as high costs in voice data collection, weakness in dynamic control, and limited intelligence. To address these challenges, this paper introduces Step-Audio, the first production-ready open-source solution. Key contribu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11946v2-abstract-full').style.display = 'inline'; document.getElementById('2502.11946v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11946v2-abstract-full" style="display: none;"> Real-time speech interaction, serving as a fundamental interface for human-machine collaboration, holds immense potential. However, current open-source models face limitations such as high costs in voice data collection, weakness in dynamic control, and limited intelligence. To address these challenges, this paper introduces Step-Audio, the first production-ready open-source solution. Key contributions include: 1) a 130B-parameter unified speech-text multi-modal model that achieves unified understanding and generation, with the Step-Audio-Chat version open-sourced; 2) a generative speech data engine that establishes an affordable voice cloning framework and produces the open-sourced lightweight Step-Audio-TTS-3B model through distillation; 3) an instruction-driven fine control system enabling dynamic adjustments across dialects, emotions, singing, and RAP; 4) an enhanced cognitive architecture augmented with tool calling and role-playing abilities to manage complex tasks effectively. Based on our new StepEval-Audio-360 evaluation benchmark, Step-Audio achieves state-of-the-art performance in human evaluations, especially in terms of instruction following. On open-source benchmarks like LLaMA Question, shows 9.3% average performance improvement, demonstrating our commitment to advancing the development of open-source multi-modal language technologies. Our code and models are available at https://github.com/stepfun-ai/Step-Audio. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11946v2-abstract-full').style.display = 'none'; document.getElementById('2502.11946v2-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 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/2502.11742">arXiv:2502.11742</a> <span> [<a href="https://arxiv.org/pdf/2502.11742">pdf</a>, <a href="https://arxiv.org/format/2502.11742">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"> Range and Bird's Eye View Fused Cross-Modal Visual Place Recognition </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Peng%2C+J">Jianyi Peng</a>, <a href="/search/cs?searchtype=author&query=Lu%2C+F">Fan Lu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin Li</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yuan Huang</a>, <a href="/search/cs?searchtype=author&query=Qu%2C+S">Sanqing Qu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+G">Guang Chen</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.11742v1-abstract-short" style="display: inline;"> Image-to-point cloud cross-modal Visual Place Recognition (VPR) is a challenging task where the query is an RGB image, and the database samples are LiDAR point clouds. Compared to single-modal VPR, this approach benefits from the widespread availability of RGB cameras and the robustness of point clouds in providing accurate spatial geometry and distance information. However, current methods rely o… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11742v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11742v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11742v1-abstract-full" style="display: none;"> Image-to-point cloud cross-modal Visual Place Recognition (VPR) is a challenging task where the query is an RGB image, and the database samples are LiDAR point clouds. Compared to single-modal VPR, this approach benefits from the widespread availability of RGB cameras and the robustness of point clouds in providing accurate spatial geometry and distance information. However, current methods rely on intermediate modalities that capture either the vertical or horizontal field of view, limiting their ability to fully exploit the complementary information from both sensors. In this work, we propose an innovative initial retrieval + re-rank method that effectively combines information from range (or RGB) images and Bird's Eye View (BEV) images. Our approach relies solely on a computationally efficient global descriptor similarity search process to achieve re-ranking. Additionally, we introduce a novel similarity label supervision technique to maximize the utility of limited training data. Specifically, we employ points average distance to approximate appearance similarity and incorporate an adaptive margin, based on similarity differences, into the vanilla triplet loss. Experimental results on the KITTI dataset demonstrate that our method significantly outperforms state-of-the-art approaches. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11742v1-abstract-full').style.display = 'none'; document.getElementById('2502.11742v1-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> <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">Submmitted to IEEE IV 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/2502.11520">arXiv:2502.11520</a> <span> [<a href="https://arxiv.org/pdf/2502.11520">pdf</a>, <a href="https://arxiv.org/format/2502.11520">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"> AURORA:Automated Training Framework of Universal Process Reward Models via Ensemble Prompting and Reverse Verification </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tan%2C+X">Xiaoyu Tan</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+T">Tianchu Yao</a>, <a href="/search/cs?searchtype=author&query=Qu%2C+C">Chao Qu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin Li</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+M">Minghao Yang</a>, <a href="/search/cs?searchtype=author&query=Lu%2C+D">Dakuan Lu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Haozhe Wang</a>, <a href="/search/cs?searchtype=author&query=Qiu%2C+X">Xihe Qiu</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+W">Wei Chu</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+Y">Yinghui Xu</a>, <a href="/search/cs?searchtype=author&query=Qi%2C+Y">Yuan Qi</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="2502.11520v1-abstract-short" style="display: inline;"> The reasoning capabilities of advanced large language models (LLMs) like o1 have revolutionized artificial intelligence applications. Nevertheless, evaluating and optimizing complex reasoning processes remain significant challenges due to diverse policy distributions and the inherent limitations of human effort and accuracy. In this paper, we present AURORA, a novel automated framework for trainin… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11520v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11520v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11520v1-abstract-full" style="display: none;"> The reasoning capabilities of advanced large language models (LLMs) like o1 have revolutionized artificial intelligence applications. Nevertheless, evaluating and optimizing complex reasoning processes remain significant challenges due to diverse policy distributions and the inherent limitations of human effort and accuracy. In this paper, we present AURORA, a novel automated framework for training universal process reward models (PRMs) using ensemble prompting and reverse verification. The framework employs a two-phase approach: First, it uses diverse prompting strategies and ensemble methods to perform automated annotation and evaluation of processes, ensuring robust assessments for reward learning. Second, it leverages practical reference answers for reverse verification, enhancing the model's ability to validate outputs and improving training accuracy. To assess the framework's performance, we extend beyond the existing ProcessBench benchmark by introducing UniversalBench, which evaluates reward predictions across full trajectories under diverse policy distribtion with long Chain-of-Thought (CoT) outputs. Experimental results demonstrate that AURORA enhances process evaluation accuracy, improves PRMs' accuracy for diverse policy distributions and long-CoT responses. The project will be open-sourced at https://auroraprm.github.io/. The Universal-PRM-7B is available at https://huggingface.co/infly/Universal-PRM-7B. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11520v1-abstract-full').style.display = 'none'; document.getElementById('2502.11520v1-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> <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</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.11279">arXiv:2502.11279</a> <span> [<a href="https://arxiv.org/pdf/2502.11279">pdf</a>, <a href="https://arxiv.org/format/2502.11279">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"> Neural Operators for Stochastic Modeling of Nonlinear Structural System Response to Natural Hazards </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Goswami%2C+S">Somdatta Goswami</a>, <a href="/search/cs?searchtype=author&query=Giovanis%2C+D+G">Dimitris G. Giovanis</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bowei Li</a>, <a href="/search/cs?searchtype=author&query=Spence%2C+S+M+J">Seymour M. J. Spence</a>, <a href="/search/cs?searchtype=author&query=Shields%2C+M+D">Michael D. Shields</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="2502.11279v1-abstract-short" style="display: inline;"> Traditionally, neural networks have been employed to learn the mapping between finite-dimensional Euclidean spaces. However, recent research has opened up new horizons, focusing on the utilization of deep neural networks to learn operators capable of mapping infinite-dimensional function spaces. In this work, we employ two state-of-the-art neural operators, the deep operator network (DeepONet) and… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11279v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11279v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11279v1-abstract-full" style="display: none;"> Traditionally, neural networks have been employed to learn the mapping between finite-dimensional Euclidean spaces. However, recent research has opened up new horizons, focusing on the utilization of deep neural networks to learn operators capable of mapping infinite-dimensional function spaces. In this work, we employ two state-of-the-art neural operators, the deep operator network (DeepONet) and the Fourier neural operator (FNO) for the prediction of the nonlinear time history response of structural systems exposed to natural hazards, such as earthquakes and wind. Specifically, we propose two architectures, a self-adaptive FNO and a Fast Fourier Transform-based DeepONet (DeepFNOnet), where we employ a FNO beyond the DeepONet to learn the discrepancy between the ground truth and the solution predicted by the DeepONet. To demonstrate the efficiency and applicability of the architectures, two problems are considered. In the first, we use the proposed model to predict the seismic nonlinear dynamic response of a six-story shear building subject to stochastic ground motions. In the second problem, we employ the operators to predict the wind-induced nonlinear dynamic response of a high-rise building while explicitly accounting for the stochastic nature of the wind excitation. In both cases, the trained metamodels achieve high accuracy while being orders of magnitude faster than their corresponding high-fidelity models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11279v1-abstract-full').style.display = 'none'; document.getElementById('2502.11279v1-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 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/2502.11079">arXiv:2502.11079</a> <span> [<a href="https://arxiv.org/pdf/2502.11079">pdf</a>, <a href="https://arxiv.org/format/2502.11079">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Phantom: Subject-consistent video generation via cross-modal alignment </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Liu%2C+L">Lijie Liu</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+T">Tianxiang Ma</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bingchuan Li</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Z">Zhuowei Chen</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jiawei Liu</a>, <a href="/search/cs?searchtype=author&query=He%2C+Q">Qian He</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+X">Xinglong Wu</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="2502.11079v1-abstract-short" style="display: inline;"> The continuous development of foundational models for video generation is evolving into various applications, with subject-consistent video generation still in the exploratory stage. We refer to this as Subject-to-Video, which extracts subject elements from reference images and generates subject-consistent video through textual instructions. We believe that the essence of subject-to-video lies in… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11079v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11079v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11079v1-abstract-full" style="display: none;"> The continuous development of foundational models for video generation is evolving into various applications, with subject-consistent video generation still in the exploratory stage. We refer to this as Subject-to-Video, which extracts subject elements from reference images and generates subject-consistent video through textual instructions. We believe that the essence of subject-to-video lies in balancing the dual-modal prompts of text and image, thereby deeply and simultaneously aligning both text and visual content. To this end, we propose Phantom, a unified video generation framework for both single and multi-subject references. Building on existing text-to-video and image-to-video architectures, we redesign the joint text-image injection model and drive it to learn cross-modal alignment via text-image-video triplet data. In particular, we emphasize subject consistency in human generation, covering existing ID-preserving video generation while offering enhanced advantages. The project homepage is here https://phantom-video.github.io/Phantom/. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11079v1-abstract-full').style.display = 'none'; document.getElementById('2502.11079v1-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 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/2502.11058">arXiv:2502.11058</a> <span> [<a href="https://arxiv.org/pdf/2502.11058">pdf</a>, <a href="https://arxiv.org/format/2502.11058">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> DreamDDP: Accelerating Data Parallel Distributed LLM Training with Layer-wise Scheduled Partial Synchronization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tang%2C+Z">Zhenheng Tang</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+Z">Zichen Tang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+J">Junlin Huang</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+X">Xinglin Pan</a>, <a href="/search/cs?searchtype=author&query=Yan%2C+R">Rudan Yan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yuxin Wang</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+A+C">Amelie Chi Zhou</a>, <a href="/search/cs?searchtype=author&query=Shi%2C+S">Shaohuai Shi</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+X">Xiaowen Chu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo 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="2502.11058v1-abstract-short" style="display: inline;"> The growth of large language models (LLMs) increases challenges of accelerating distributed training across multiple GPUs in different data centers. Moreover, concerns about data privacy and data exhaustion have heightened interest in geo-distributed data centers. Communication in geo-distributed data parallel training (DDP) with stochastic gradient descent (S-SGD) is the main bottleneck in low-ba… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11058v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11058v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11058v1-abstract-full" style="display: none;"> The growth of large language models (LLMs) increases challenges of accelerating distributed training across multiple GPUs in different data centers. Moreover, concerns about data privacy and data exhaustion have heightened interest in geo-distributed data centers. Communication in geo-distributed data parallel training (DDP) with stochastic gradient descent (S-SGD) is the main bottleneck in low-bandwidth environments. Local SGD mitigates communication overhead by reducing synchronization frequency, and recent studies have successfully applied it to geo-distributedly pre-train LLMs. However, we identify that its model synchronization mechanism prevents overlapping communication and computation, which makes the system lose opportunities to overlap communication and computation. To overcome this limitation, we expand the design space of local SGD by layer-wisely decoupling model synchronization. In each iteration, only some layers are synchronized instead of the entire model after a specific number of iterations. Leveraging this methodology, we introduce DreamDDP, a training framework to accelerate low-bandwidth distributed training with three key innovations: (1) partial local SGD with theoretical assurances of convergence rates comparable to S-SGD; (2) overlapping parameter synchronization with computation without extra GPU memory occupation; (3) identifying and exploiting three properties to schedule the communication and computation to reduce the training time based on fine-grained profiling of layer-wise communication and computation time. Empirical evaluations conducted on 32 GPUs using prominent deep learning models, including ResNet-18, ResNet-50, GPT-2, and Llama-2, demonstrate that DreamDDP enhances the convergence properties of Local SGD (and Adam) and achieves speedups ranging from $1.49\times$ to $3.91\times$ over leading baseline methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11058v1-abstract-full').style.display = 'none'; document.getElementById('2502.11058v1-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 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/2502.10761">arXiv:2502.10761</a> <span> [<a href="https://arxiv.org/pdf/2502.10761">pdf</a>, <a href="https://arxiv.org/format/2502.10761">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 Whole-Body Disturbance Rejection Control Framework for Dynamic Motions in Legged Robots </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Bolin Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wentao Zhang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+X">Xuecong Huang</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+L">Lijun Zhu</a>, <a href="/search/cs?searchtype=author&query=Ding%2C+H">Han Ding</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="2502.10761v2-abstract-short" style="display: inline;"> This letter presents a control framework for legged robots that enables self-perception and resistance to external disturbances and model uncertainties. First, a novel disturbance estimator is proposed, integrating adaptive control and extended state observers (ESO) to estimate external disturbances and model uncertainties. This estimator is embedded within the whole-body control framework to comp… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10761v2-abstract-full').style.display = 'inline'; document.getElementById('2502.10761v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10761v2-abstract-full" style="display: none;"> This letter presents a control framework for legged robots that enables self-perception and resistance to external disturbances and model uncertainties. First, a novel disturbance estimator is proposed, integrating adaptive control and extended state observers (ESO) to estimate external disturbances and model uncertainties. This estimator is embedded within the whole-body control framework to compensate for disturbances in the legged system. Second, a comprehensive whole-body disturbance rejection control framework (WB-DRC) is introduced, accounting for the robot's full-body dynamics. Compared to previous whole-body control frameworks, WB-DRC effectively handles external disturbances and model uncertainties, with the potential to adapt to complex terrain. Third, simulations of both biped and quadruped robots are conducted in the Gazebo simulator to demonstrate the effectiveness and versatility of WB-DRC. Finally, extensive experimental trials on the quadruped robot validate the robustness and stability of the robot system using WB-DRC under various disturbance conditions. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10761v2-abstract-full').style.display = 'none'; document.getElementById('2502.10761v2-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">v1</span> submitted 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">have submitted to IEEE RA-L</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.10631">arXiv:2502.10631</a> <span> [<a href="https://arxiv.org/pdf/2502.10631">pdf</a>, <a href="https://arxiv.org/format/2502.10631">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="Biomolecules">q-bio.BM</span> </div> </div> <p class="title is-5 mathjax"> ControllableGPT: A Ground-Up Designed Controllable GPT for Molecule Optimization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xuefeng Liu</a>, <a href="/search/cs?searchtype=author&query=Jiang%2C+S">Songhao Jiang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Stevens%2C+R">Rick Stevens</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="2502.10631v1-abstract-short" style="display: inline;"> Large Language Models (LLMs) employ three popular training approaches: Masked Language Models (MLM), Causal Language Models (CLM), and Sequence-to-Sequence Models (seq2seq). However, each approach has its strengths and limitations, and faces challenges in addressing specific tasks that require controllable and bidirectional generation, such as drug optimization. To address this challenge, inspired… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10631v1-abstract-full').style.display = 'inline'; document.getElementById('2502.10631v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10631v1-abstract-full" style="display: none;"> Large Language Models (LLMs) employ three popular training approaches: Masked Language Models (MLM), Causal Language Models (CLM), and Sequence-to-Sequence Models (seq2seq). However, each approach has its strengths and limitations, and faces challenges in addressing specific tasks that require controllable and bidirectional generation, such as drug optimization. To address this challenge, inspired by the biological processes of growth and evolution, which involve the expansion, shrinking, and mutation of sequences, we introduce ControllableGPT. This initiative represents the first effort to combine the advantages of MLM, CLM, and seq2seq into a single unified, controllable GPT framework. It enables the precise management of specific locations and ranges within a sequence, allowing for expansion, reduction, or mutation over chosen or random lengths, while maintaining the integrity of any specified positions or subsequences. In this work, we designed ControllableGPT for drug optimization from the ground up, which included proposing the Causally Masked Seq2seq (CMS) objective, developing the training corpus, introducing a novel pre-training approach, and devising a unique generation process. We demonstrate the effectiveness and controllability of ControllableGPT by conducting experiments on drug optimization tasks for both viral and cancer benchmarks, surpassing competing baselines. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10631v1-abstract-full').style.display = 'none'; document.getElementById('2502.10631v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.10248">arXiv:2502.10248</a> <span> [<a href="https://arxiv.org/pdf/2502.10248">pdf</a>, <a href="https://arxiv.org/format/2502.10248">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="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ma%2C+G">Guoqing Ma</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+H">Haoyang Huang</a>, <a href="/search/cs?searchtype=author&query=Yan%2C+K">Kun Yan</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+L">Liangyu Chen</a>, <a href="/search/cs?searchtype=author&query=Duan%2C+N">Nan Duan</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+S">Shengming Yin</a>, <a href="/search/cs?searchtype=author&query=Wan%2C+C">Changyi Wan</a>, <a href="/search/cs?searchtype=author&query=Ming%2C+R">Ranchen Ming</a>, <a href="/search/cs?searchtype=author&query=Song%2C+X">Xiaoniu Song</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xing Chen</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Y">Yu Zhou</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+D">Deshan Sun</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+D">Deyu Zhou</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+J">Jian Zhou</a>, <a href="/search/cs?searchtype=author&query=Tan%2C+K">Kaijun Tan</a>, <a href="/search/cs?searchtype=author&query=An%2C+K">Kang An</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+M">Mei Chen</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+W">Wei Ji</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+Q">Qiling Wu</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+W">Wen Sun</a>, <a href="/search/cs?searchtype=author&query=Han%2C+X">Xin Han</a>, <a href="/search/cs?searchtype=author&query=Wei%2C+Y">Yanan Wei</a>, <a href="/search/cs?searchtype=author&query=Ge%2C+Z">Zheng Ge</a>, <a href="/search/cs?searchtype=author&query=Li%2C+A">Aojie Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+B">Bin Wang</a> , et al. (90 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="2502.10248v3-abstract-short" style="display: inline;"> We present Step-Video-T2V, a state-of-the-art text-to-video pre-trained model with 30B parameters and the ability to generate videos up to 204 frames in length. A deep compression Variational Autoencoder, Video-VAE, is designed for video generation tasks, achieving 16x16 spatial and 8x temporal compression ratios, while maintaining exceptional video reconstruction quality. User prompts are encoded… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10248v3-abstract-full').style.display = 'inline'; document.getElementById('2502.10248v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10248v3-abstract-full" style="display: none;"> We present Step-Video-T2V, a state-of-the-art text-to-video pre-trained model with 30B parameters and the ability to generate videos up to 204 frames in length. A deep compression Variational Autoencoder, Video-VAE, is designed for video generation tasks, achieving 16x16 spatial and 8x temporal compression ratios, while maintaining exceptional video reconstruction quality. User prompts are encoded using two bilingual text encoders to handle both English and Chinese. A DiT with 3D full attention is trained using Flow Matching and is employed to denoise input noise into latent frames. A video-based DPO approach, Video-DPO, is applied to reduce artifacts and improve the visual quality of the generated videos. We also detail our training strategies and share key observations and insights. Step-Video-T2V's performance is evaluated on a novel video generation benchmark, Step-Video-T2V-Eval, demonstrating its state-of-the-art text-to-video quality when compared with both open-source and commercial engines. Additionally, we discuss the limitations of current diffusion-based model paradigm and outline future directions for video foundation models. We make both Step-Video-T2V and Step-Video-T2V-Eval available at https://github.com/stepfun-ai/Step-Video-T2V. The online version can be accessed from https://yuewen.cn/videos as well. Our goal is to accelerate the innovation of video foundation models and empower video content creators. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10248v3-abstract-full').style.display = 'none'; document.getElementById('2502.10248v3-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">v1</span> submitted 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">36 pages, 14 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/2502.10047">arXiv:2502.10047</a> <span> [<a href="https://arxiv.org/pdf/2502.10047">pdf</a>, <a href="https://arxiv.org/format/2502.10047">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Janus: Collaborative Vision Transformer Under Dynamic Network Environment </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jiang%2C+L">Linyi Jiang</a>, <a href="/search/cs?searchtype=author&query=Fu%2C+S+D">Silvery D. Fu</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+Y">Yifei Zhu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo 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="2502.10047v1-abstract-short" style="display: inline;"> Vision Transformers (ViTs) have outperformed traditional Convolutional Neural Network architectures and achieved state-of-the-art results in various computer vision tasks. Since ViTs are computationally expensive, the models either have to be pruned to run on resource-limited edge devices only or have to be executed on remote cloud servers after receiving the raw data transmitted over fluctuating… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10047v1-abstract-full').style.display = 'inline'; document.getElementById('2502.10047v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10047v1-abstract-full" style="display: none;"> Vision Transformers (ViTs) have outperformed traditional Convolutional Neural Network architectures and achieved state-of-the-art results in various computer vision tasks. Since ViTs are computationally expensive, the models either have to be pruned to run on resource-limited edge devices only or have to be executed on remote cloud servers after receiving the raw data transmitted over fluctuating networks. The resulting degraded performance or high latency all hinder their widespread applications. In this paper, we present Janus, the first framework for low-latency cloud-device collaborative Vision Transformer inference over dynamic networks. Janus overcomes the intrinsic model limitations of ViTs and realizes collaboratively executing ViT models on both cloud and edge devices, achieving low latency, high accuracy, and low communication overhead. Specifically, Janus judiciously combines token pruning techniques with a carefully designed fine-to-coarse model splitting policy and non-static mixed pruning policy. It attains a balance between accuracy and latency by dynamically selecting the optimal pruning level and split point. Experimental results across various tasks demonstrate that Janus enhances throughput by up to 5.15 times and reduces latency violation ratios by up to 98.7% when compared with baseline approaches under various network environments. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10047v1-abstract-full').style.display = 'none'; document.getElementById('2502.10047v1-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">Accepted for publication in IEEE INFOCOM 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/2502.08916">arXiv:2502.08916</a> <span> [<a href="https://arxiv.org/pdf/2502.08916">pdf</a>, <a href="https://arxiv.org/format/2502.08916">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="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Multiagent Systems">cs.MA</span> </div> </div> <p class="title is-5 mathjax"> PathFinder: A Multi-Modal Multi-Agent System for Medical Diagnostic Decision-Making Applied to Histopathology </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ghezloo%2C+F">Fatemeh Ghezloo</a>, <a href="/search/cs?searchtype=author&query=Seyfioglu%2C+M+S">Mehmet Saygin Seyfioglu</a>, <a href="/search/cs?searchtype=author&query=Soraki%2C+R">Rustin Soraki</a>, <a href="/search/cs?searchtype=author&query=Ikezogwo%2C+W+O">Wisdom O. Ikezogwo</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Beibin Li</a>, <a href="/search/cs?searchtype=author&query=Vivekanandan%2C+T">Tejoram Vivekanandan</a>, <a href="/search/cs?searchtype=author&query=Elmore%2C+J+G">Joann G. Elmore</a>, <a href="/search/cs?searchtype=author&query=Krishna%2C+R">Ranjay Krishna</a>, <a href="/search/cs?searchtype=author&query=Shapiro%2C+L">Linda Shapiro</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="2502.08916v1-abstract-short" style="display: inline;"> Diagnosing diseases through histopathology whole slide images (WSIs) is fundamental in modern pathology but is challenged by the gigapixel scale and complexity of WSIs. Trained histopathologists overcome this challenge by navigating the WSI, looking for relevant patches, taking notes, and compiling them to produce a final holistic diagnostic. Traditional AI approaches, such as multiple instance le… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08916v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08916v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08916v1-abstract-full" style="display: none;"> Diagnosing diseases through histopathology whole slide images (WSIs) is fundamental in modern pathology but is challenged by the gigapixel scale and complexity of WSIs. Trained histopathologists overcome this challenge by navigating the WSI, looking for relevant patches, taking notes, and compiling them to produce a final holistic diagnostic. Traditional AI approaches, such as multiple instance learning and transformer-based models, fail short of such a holistic, iterative, multi-scale diagnostic procedure, limiting their adoption in the real-world. We introduce PathFinder, a multi-modal, multi-agent framework that emulates the decision-making process of expert pathologists. PathFinder integrates four AI agents, the Triage Agent, Navigation Agent, Description Agent, and Diagnosis Agent, that collaboratively navigate WSIs, gather evidence, and provide comprehensive diagnoses with natural language explanations. The Triage Agent classifies the WSI as benign or risky; if risky, the Navigation and Description Agents iteratively focus on significant regions, generating importance maps and descriptive insights of sampled patches. Finally, the Diagnosis Agent synthesizes the findings to determine the patient's diagnostic classification. Our Experiments show that PathFinder outperforms state-of-the-art methods in skin melanoma diagnosis by 8% while offering inherent explainability through natural language descriptions of diagnostically relevant patches. Qualitative analysis by pathologists shows that the Description Agent's outputs are of high quality and comparable to GPT-4o. PathFinder is also the first AI-based system to surpass the average performance of pathologists in this challenging melanoma classification task by 9%, setting a new record for efficient, accurate, and interpretable AI-assisted diagnostics in pathology. Data, code and models available at https://pathfinder-dx.github.io/ <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08916v1-abstract-full').style.display = 'none'; document.getElementById('2502.08916v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.08697">arXiv:2502.08697</a> <span> [<a href="https://arxiv.org/pdf/2502.08697">pdf</a>, <a href="https://arxiv.org/format/2502.08697">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"> Bilevel Learning for Bilevel Planning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+B">Bowen Li</a>, <a href="/search/cs?searchtype=author&query=Silver%2C+T">Tom Silver</a>, <a href="/search/cs?searchtype=author&query=Scherer%2C+S">Sebastian Scherer</a>, <a href="/search/cs?searchtype=author&query=Gray%2C+A">Alexander Gray</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="2502.08697v2-abstract-short" style="display: inline;"> A robot that learns from demonstrations should not just imitate what it sees -- it should understand the high-level concepts that are being demonstrated and generalize them to new tasks. Bilevel planning is a hierarchical model-based approach where predicates (relational state abstractions) can be leveraged to achieve compositional generalization. However, previous bilevel planning approaches depe… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08697v2-abstract-full').style.display = 'inline'; document.getElementById('2502.08697v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08697v2-abstract-full" style="display: none;"> A robot that learns from demonstrations should not just imitate what it sees -- it should understand the high-level concepts that are being demonstrated and generalize them to new tasks. Bilevel planning is a hierarchical model-based approach where predicates (relational state abstractions) can be leveraged to achieve compositional generalization. However, previous bilevel planning approaches depend on predicates that are either hand-engineered or restricted to very simple forms, limiting their scalability to sophisticated, high-dimensional state spaces. To address this limitation, we present IVNTR, the first bilevel planning approach capable of learning neural predicates directly from demonstrations. Our key innovation is a neuro-symbolic bilevel learning framework that mirrors the structure of bilevel planning. In IVNTR, symbolic learning of the predicate "effects" and neural learning of the predicate "functions" alternate, with each providing guidance for the other. We evaluate IVNTR in six diverse robot planning domains, demonstrating its effectiveness in abstracting various continuous and high-dimensional states. While most existing approaches struggle to generalize (with <35% success rate), our IVNTR achieves an average of 77% success rate on unseen tasks. Additionally, we showcase IVNTR on a mobile manipulator, where it learns to perform real-world mobile manipulation tasks and generalizes to unseen test scenarios that feature new objects, new states, and longer task horizons. Our findings underscore the promise of learning and planning with abstractions as a path towards high-level generalization. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08697v2-abstract-full').style.display = 'none'; document.getElementById('2502.08697v2-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 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 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">16 pages, 9 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/2502.06808">arXiv:2502.06808</a> <span> [<a href="https://arxiv.org/pdf/2502.06808">pdf</a>, <a href="https://arxiv.org/format/2502.06808">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"> On the Benefits of Attribute-Driven Graph Domain Adaptation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Fang%2C+R">Ruiyi Fang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bingheng Li</a>, <a href="/search/cs?searchtype=author&query=Kang%2C+Z">Zhao Kang</a>, <a href="/search/cs?searchtype=author&query=Zeng%2C+Q">Qiuhao Zeng</a>, <a href="/search/cs?searchtype=author&query=Dashtbayaz%2C+N+H">Nima Hosseini Dashtbayaz</a>, <a href="/search/cs?searchtype=author&query=Pu%2C+R">Ruizhi Pu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+B">Boyu Wang</a>, <a href="/search/cs?searchtype=author&query=Ling%2C+C">Charles Ling</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="2502.06808v3-abstract-short" style="display: inline;"> Graph Domain Adaptation (GDA) addresses a pressing challenge in cross-network learning, particularly pertinent due to the absence of labeled data in real-world graph datasets. Recent studies attempted to learn domain invariant representations by eliminating structural shifts between graphs. In this work, we show that existing methodologies have overlooked the significance of the graph node attribu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06808v3-abstract-full').style.display = 'inline'; document.getElementById('2502.06808v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06808v3-abstract-full" style="display: none;"> Graph Domain Adaptation (GDA) addresses a pressing challenge in cross-network learning, particularly pertinent due to the absence of labeled data in real-world graph datasets. Recent studies attempted to learn domain invariant representations by eliminating structural shifts between graphs. In this work, we show that existing methodologies have overlooked the significance of the graph node attribute, a pivotal factor for graph domain alignment. Specifically, we first reveal the impact of node attributes for GDA by theoretically proving that in addition to the graph structural divergence between the domains, the node attribute discrepancy also plays a critical role in GDA. Moreover, we also empirically show that the attribute shift is more substantial than the topology shift, which further underscores the importance of node attribute alignment in GDA. Inspired by this finding, a novel cross-channel module is developed to fuse and align both views between the source and target graphs for GDA. Experimental results on a variety of benchmarks verify the effectiveness of our method. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06808v3-abstract-full').style.display = 'none'; document.getElementById('2502.06808v3-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">v1</span> submitted 3 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">Accepted by the ICLR 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/2502.06773">arXiv:2502.06773</a> <span> [<a href="https://arxiv.org/pdf/2502.06773">pdf</a>, <a href="https://arxiv.org/format/2502.06773">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> On the Emergence of Thinking in LLMs I: Searching for the Right Intuition </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ye%2C+G">Guanghao Ye</a>, <a href="/search/cs?searchtype=author&query=Pham%2C+K+D">Khiem Duc Pham</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+X">Xinzhi Zhang</a>, <a href="/search/cs?searchtype=author&query=Gopi%2C+S">Sivakanth Gopi</a>, <a href="/search/cs?searchtype=author&query=Peng%2C+B">Baolin Peng</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Beibin Li</a>, <a href="/search/cs?searchtype=author&query=Kulkarni%2C+J">Janardhan Kulkarni</a>, <a href="/search/cs?searchtype=author&query=Inan%2C+H+A">Huseyin A. Inan</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="2502.06773v1-abstract-short" style="display: inline;"> Recent AI advancements, such as OpenAI's new models, are transforming LLMs into LRMs (Large Reasoning Models) that perform reasoning during inference, taking extra time and compute for higher-quality outputs. We aim to uncover the algorithmic framework for training LRMs. Methods like self-consistency, PRM, and AlphaZero suggest reasoning as guided search. We ask: what is the simplest, most scalabl… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06773v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06773v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06773v1-abstract-full" style="display: none;"> Recent AI advancements, such as OpenAI's new models, are transforming LLMs into LRMs (Large Reasoning Models) that perform reasoning during inference, taking extra time and compute for higher-quality outputs. We aim to uncover the algorithmic framework for training LRMs. Methods like self-consistency, PRM, and AlphaZero suggest reasoning as guided search. We ask: what is the simplest, most scalable way to enable search in LLMs? We propose a post-training framework called Reinforcement Learning via Self-Play (RLSP). RLSP involves three steps: (1) supervised fine-tuning with human or synthetic demonstrations of the reasoning process, (2) using an exploration reward signal to encourage diverse and efficient reasoning behaviors, and (3) RL training with an outcome verifier to ensure correctness while preventing reward hacking. Our key innovation is to decouple exploration and correctness signals during PPO training, carefully balancing them to improve performance and efficiency. Empirical studies in the math domain show that RLSP improves reasoning. On the Llama-3.1-8B-Instruct model, RLSP can boost performance by 23% in MATH-500 test set; On AIME 2024 math problems, Qwen2.5-32B-Instruct improved by 10% due to RLSP. However, a more important finding of this work is that the models trained using RLSP, even with the simplest exploration reward that encourages the model to take more intermediate steps, showed several emergent behaviors such as backtracking, exploration of ideas, and verification. These findings demonstrate that RLSP framework might be enough to enable emergence of complex reasoning abilities in LLMs when scaled. Lastly, we propose a theory as to why RLSP search strategy is more suitable for LLMs inspired by a remarkable result that says CoT provably increases computational power of LLMs, which grows as the number of steps in CoT \cite{li2024chain,merrill2023expresssive}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06773v1-abstract-full').style.display = 'none'; document.getElementById('2502.06773v1-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">Abstract shortened for arXiv</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06563">arXiv:2502.06563</a> <span> [<a href="https://arxiv.org/pdf/2502.06563">pdf</a>, <a href="https://arxiv.org/format/2502.06563">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"> Large Language Models Meet Symbolic Provers for Logical Reasoning Evaluation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Qi%2C+C">Chengwen Qi</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+R">Ren Ma</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bowen Li</a>, <a href="/search/cs?searchtype=author&query=Du%2C+H">He Du</a>, <a href="/search/cs?searchtype=author&query=Hui%2C+B">Binyuan Hui</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+J">Jinwang Wu</a>, <a href="/search/cs?searchtype=author&query=Laili%2C+Y">Yuanjun Laili</a>, <a href="/search/cs?searchtype=author&query=He%2C+C">Conghui He</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="2502.06563v1-abstract-short" style="display: inline;"> First-order logic (FOL) reasoning, which involves sequential deduction, is pivotal for intelligent systems and serves as a valuable task for evaluating reasoning capabilities, particularly in chain-of-thought (CoT) contexts. Existing benchmarks often rely on extensive human annotation or handcrafted templates, making it difficult to achieve the necessary complexity, scalability, and diversity for… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06563v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06563v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06563v1-abstract-full" style="display: none;"> First-order logic (FOL) reasoning, which involves sequential deduction, is pivotal for intelligent systems and serves as a valuable task for evaluating reasoning capabilities, particularly in chain-of-thought (CoT) contexts. Existing benchmarks often rely on extensive human annotation or handcrafted templates, making it difficult to achieve the necessary complexity, scalability, and diversity for robust evaluation. To address these limitations, we propose a novel framework called ProverGen that synergizes the generative strengths of Large Language Models (LLMs) with the rigor and precision of symbolic provers, enabling the creation of a scalable, diverse, and high-quality FOL reasoning dataset, ProverQA. ProverQA is also distinguished by its inclusion of accessible and logically coherent intermediate reasoning steps for each problem. Our evaluation shows that state-of-the-art LLMs struggle to solve ProverQA problems, even with CoT prompting, highlighting the dataset's challenging nature. We also finetune Llama3.1-8B-Instruct on a separate training set generated by our framework. The finetuned model demonstrates consistent improvements on both in-distribution and out-of-distribution test sets, suggesting the value of our proposed data generation framework. Code available at: https://github.com/opendatalab/ProverGen <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06563v1-abstract-full').style.display = 'none'; document.getElementById('2502.06563v1-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">Accepted by ICLR 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/2502.06490">arXiv:2502.06490</a> <span> [<a href="https://arxiv.org/pdf/2502.06490">pdf</a>, <a href="https://arxiv.org/format/2502.06490">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</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="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="Signal Processing">eess.SP</span> </div> </div> <p class="title is-5 mathjax"> Recent Advances in Discrete Speech Tokens: A Review </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Guo%2C+Y">Yiwei Guo</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Z">Zhihan Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Hankun Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bohan Li</a>, <a href="/search/cs?searchtype=author&query=Shao%2C+C">Chongtian Shao</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Hanglei Zhang</a>, <a href="/search/cs?searchtype=author&query=Du%2C+C">Chenpeng Du</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xie Chen</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+S">Shujie Liu</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+K">Kai Yu</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="2502.06490v2-abstract-short" style="display: inline;"> The rapid advancement of speech generation technologies in the era of large language models (LLMs) has established discrete speech tokens as a foundational paradigm for speech representation. These tokens, characterized by their discrete, compact, and concise nature, are not only advantageous for efficient transmission and storage, but also inherently compatible with the language modeling framewor… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06490v2-abstract-full').style.display = 'inline'; document.getElementById('2502.06490v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06490v2-abstract-full" style="display: none;"> The rapid advancement of speech generation technologies in the era of large language models (LLMs) has established discrete speech tokens as a foundational paradigm for speech representation. These tokens, characterized by their discrete, compact, and concise nature, are not only advantageous for efficient transmission and storage, but also inherently compatible with the language modeling framework, enabling seamless integration of speech into text-dominated LLM architectures. Current research categorizes discrete speech tokens into two principal classes: acoustic tokens and semantic tokens, each of which has evolved into a rich research domain characterized by unique design philosophies and methodological approaches. This survey systematically synthesizes the existing taxonomy and recent innovations in discrete speech tokenization, conducts a critical examination of the strengths and limitations of each paradigm, and presents systematic experimental comparisons across token types. Furthermore, we identify persistent challenges in the field and propose potential research directions, aiming to offer actionable insights to inspire future advancements in the development and application of discrete speech tokens. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06490v2-abstract-full').style.display = 'none'; document.getElementById('2502.06490v2-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 February, 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">23 pages, 8 figures, 3 tables. Work in progress</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.05928">arXiv:2502.05928</a> <span> [<a href="https://arxiv.org/pdf/2502.05928">pdf</a>, <a href="https://arxiv.org/format/2502.05928">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"> ClinKD: Cross-Modal Clinic Knowledge Distiller For Multi-Task Medical Images </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ge%2C+H">Hongyu Ge</a>, <a href="/search/cs?searchtype=author&query=Hao%2C+L">Longkun Hao</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+Z">Zihui Xu</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+Z">Zhenxin Lin</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin Li</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+S">Shoujun Zhou</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+H">Hongjin Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Y">Yihang Liu</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.05928v1-abstract-short" style="display: inline;"> Med-VQA (Medical Visual Question Answering) is a crucial subtask within the broader VQA (Visual Question Answering) domain. This task requires a visual question answering system to analyze the provided image and corresponding question,offering reasonable analysis and suggestions to assist medical professionals in making pathological diagnoses, or ideally, enabling the system to independently provi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05928v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05928v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05928v1-abstract-full" style="display: none;"> Med-VQA (Medical Visual Question Answering) is a crucial subtask within the broader VQA (Visual Question Answering) domain. This task requires a visual question answering system to analyze the provided image and corresponding question,offering reasonable analysis and suggestions to assist medical professionals in making pathological diagnoses, or ideally, enabling the system to independently provide correct diagnoses. Furthermore, more advanced Med-VQA tasks involve Referring and Grounding, which not only require the system to accurately comprehend medical images but also to pinpoint specific biological locations within those images. While many large pre-trained models have demonstrated substantial VQA capabilities,challenges persist in the medical imaging domain. The intricacy of biological features in medical images and the scarcity of high-quality medical image datasets, combined with the fact that current models are not tailored for the medical field in terms of architecture and training paradigms, hinder the full exploitation of model generalization. This results in issues such as hallucination in Visual Grounding. In this paper, we introduce the ClinKD model, which incorporates modifications to model position encoding and a diversified training process. Initially, we enhance the model's ability to perceive image and modality variations by using Med-CLIP Guided Rotary Position Embedding. Subsequently, we leverage distillation to provide prior knowledge to the model before using complete training data. Additionally, the feedback-based training process during the formal training phase further enhances data utilization. Notably, under unchanged evaluation protocols, we achieve a new state-of-the-art performance on the Med-GRIT-270k dataset, and the Med-CLIP Guided Rotary Position Embedding approach presents potential for generalizing to universal model position encoding. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05928v1-abstract-full').style.display = 'none'; document.getElementById('2502.05928v1-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/2502.05859">arXiv:2502.05859</a> <span> [<a href="https://arxiv.org/pdf/2502.05859">pdf</a>, <a href="https://arxiv.org/format/2502.05859">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"> SphereFusion: Efficient Panorama Depth Estimation via Gated Fusion </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yan%2C+Q">Qingsong Yan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Q">Qiang Wang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+K">Kaiyong Zhao</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+J">Jie Chen</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+X">Xiaowen Chu</a>, <a href="/search/cs?searchtype=author&query=Deng%2C+F">Fei Deng</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="2502.05859v1-abstract-short" style="display: inline;"> Due to the rapid development of panorama cameras, the task of estimating panorama depth has attracted significant attention from the computer vision community, especially in applications such as robot sensing and autonomous driving. However, existing methods relying on different projection formats often encounter challenges, either struggling with distortion and discontinuity in the case of equire… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05859v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05859v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05859v1-abstract-full" style="display: none;"> Due to the rapid development of panorama cameras, the task of estimating panorama depth has attracted significant attention from the computer vision community, especially in applications such as robot sensing and autonomous driving. However, existing methods relying on different projection formats often encounter challenges, either struggling with distortion and discontinuity in the case of equirectangular, cubemap, and tangent projections, or experiencing a loss of texture details with the spherical projection. To tackle these concerns, we present SphereFusion, an end-to-end framework that combines the strengths of various projection methods. Specifically, SphereFusion initially employs 2D image convolution and mesh operations to extract two distinct types of features from the panorama image in both equirectangular and spherical projection domains. These features are then projected onto the spherical domain, where a gate fusion module selects the most reliable features for fusion. Finally, SphereFusion estimates panorama depth within the spherical domain. Meanwhile, SphereFusion employs a cache strategy to improve the efficiency of mesh operation. Extensive experiments on three public panorama datasets demonstrate that SphereFusion achieves competitive results with other state-of-the-art methods, while presenting the fastest inference speed at only 17 ms on a 512$\times$1024 panorama image. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05859v1-abstract-full').style.display = 'none'; document.getElementById('2502.05859v1-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> <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">3DV 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/2502.05534">arXiv:2502.05534</a> <span> [<a href="https://arxiv.org/pdf/2502.05534">pdf</a>, <a href="https://arxiv.org/format/2502.05534">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"> Fg-T2M++: LLMs-Augmented Fine-Grained Text Driven Human Motion Generation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yin Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+M">Mu Li</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+J">Jiapeng Liu</a>, <a href="/search/cs?searchtype=author&query=Leng%2C+Z">Zhiying Leng</a>, <a href="/search/cs?searchtype=author&query=Li%2C+F+W+B">Frederick W. B. Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Ziyao Zhang</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+X">Xiaohui Liang</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="2502.05534v1-abstract-short" style="display: inline;"> We address the challenging problem of fine-grained text-driven human motion generation. Existing works generate imprecise motions that fail to accurately capture relationships specified in text due to: (1) lack of effective text parsing for detailed semantic cues regarding body parts, (2) not fully modeling linguistic structures between words to comprehend text comprehensively. To tackle these lim… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05534v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05534v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05534v1-abstract-full" style="display: none;"> We address the challenging problem of fine-grained text-driven human motion generation. Existing works generate imprecise motions that fail to accurately capture relationships specified in text due to: (1) lack of effective text parsing for detailed semantic cues regarding body parts, (2) not fully modeling linguistic structures between words to comprehend text comprehensively. To tackle these limitations, we propose a novel fine-grained framework Fg-T2M++ that consists of: (1) an LLMs semantic parsing module to extract body part descriptions and semantics from text, (2) a hyperbolic text representation module to encode relational information between text units by embedding the syntactic dependency graph into hyperbolic space, and (3) a multi-modal fusion module to hierarchically fuse text and motion features. Extensive experiments on HumanML3D and KIT-ML datasets demonstrate that Fg-T2M++ outperforms SOTA methods, validating its ability to accurately generate motions adhering to comprehensive text semantics. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05534v1-abstract-full').style.display = 'none'; document.getElementById('2502.05534v1-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 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/2502.05206">arXiv:2502.05206</a> <span> [<a href="https://arxiv.org/pdf/2502.05206">pdf</a>, <a href="https://arxiv.org/format/2502.05206">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> Safety at Scale: A Comprehensive Survey of Large Model Safety </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ma%2C+X">Xingjun Ma</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+Y">Yifeng Gao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yixu Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruofan Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xin Wang</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+Y">Ye Sun</a>, <a href="/search/cs?searchtype=author&query=Ding%2C+Y">Yifan Ding</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+H">Hengyuan Xu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yunhao Chen</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+Y">Yunhan Zhao</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+H">Hanxun Huang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yige Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jiaming Zhang</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+X">Xiang Zheng</a>, <a href="/search/cs?searchtype=author&query=Bai%2C+Y">Yang Bai</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+Z">Zuxuan Wu</a>, <a href="/search/cs?searchtype=author&query=Qiu%2C+X">Xipeng Qiu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jingfeng Zhang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yiming Li</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+J">Jun Sun</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+C">Cong Wang</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+J">Jindong Gu</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+B">Baoyuan Wu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+S">Siheng Chen</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+T">Tianwei Zhang</a> , et al. (19 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="2502.05206v2-abstract-short" style="display: inline;"> The rapid advancement of large models, driven by their exceptional abilities in learning and generalization through large-scale pre-training, has reshaped the landscape of Artificial Intelligence (AI). These models are now foundational to a wide range of applications, including conversational AI, recommendation systems, autonomous driving, content generation, medical diagnostics, and scientific di… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05206v2-abstract-full').style.display = 'inline'; document.getElementById('2502.05206v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05206v2-abstract-full" style="display: none;"> The rapid advancement of large models, driven by their exceptional abilities in learning and generalization through large-scale pre-training, has reshaped the landscape of Artificial Intelligence (AI). These models are now foundational to a wide range of applications, including conversational AI, recommendation systems, autonomous driving, content generation, medical diagnostics, and scientific discovery. However, their widespread deployment also exposes them to significant safety risks, raising concerns about robustness, reliability, and ethical implications. This survey provides a systematic review of current safety research on large models, covering Vision Foundation Models (VFMs), Large Language Models (LLMs), Vision-Language Pre-training (VLP) models, Vision-Language Models (VLMs), Diffusion Models (DMs), and large-model-based Agents. Our contributions are summarized as follows: (1) We present a comprehensive taxonomy of safety threats to these models, including adversarial attacks, data poisoning, backdoor attacks, jailbreak and prompt injection attacks, energy-latency attacks, data and model extraction attacks, and emerging agent-specific threats. (2) We review defense strategies proposed for each type of attacks if available and summarize the commonly used datasets and benchmarks for safety research. (3) Building on this, we identify and discuss the open challenges in large model safety, emphasizing the need for comprehensive safety evaluations, scalable and effective defense mechanisms, and sustainable data practices. More importantly, we highlight the necessity of collective efforts from the research community and international collaboration. Our work can serve as a useful reference for researchers and practitioners, fostering the ongoing development of comprehensive defense systems and platforms to safeguard AI models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05206v2-abstract-full').style.display = 'none'; document.getElementById('2502.05206v2-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">v1</span> submitted 2 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">47 pages, 3 figures, 11 tables GitHub: https://github.com/xingjunm/Awesome-Large-Model-Safety</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.05163">arXiv:2502.05163</a> <span> [<a href="https://arxiv.org/pdf/2502.05163">pdf</a>, <a href="https://arxiv.org/format/2502.05163">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> DuoGuard: A Two-Player RL-Driven Framework for Multilingual LLM Guardrails </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Deng%2C+Y">Yihe Deng</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Y">Yu Yang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Junkai Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+W">Wei Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo 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="2502.05163v1-abstract-short" style="display: inline;"> The rapid advancement of large language models (LLMs) has increased the need for guardrail models to ensure responsible use, particularly in detecting unsafe and illegal content. While substantial safety data exist in English, multilingual guardrail modeling remains underexplored due to the scarcity of open-source safety data in other languages. To address this gap, we propose a novel two-player R… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05163v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05163v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05163v1-abstract-full" style="display: none;"> The rapid advancement of large language models (LLMs) has increased the need for guardrail models to ensure responsible use, particularly in detecting unsafe and illegal content. While substantial safety data exist in English, multilingual guardrail modeling remains underexplored due to the scarcity of open-source safety data in other languages. To address this gap, we propose a novel two-player Reinforcement Learning (RL) framework, where a generator and a guardrail model co-evolve adversarially to produce high-quality synthetic data for multilingual guardrail training. We theoretically formalize this interaction as a two-player game, proving convergence to a Nash equilibrium. Empirical evaluations show that our model \ours outperforms state-of-the-art models, achieving nearly 10% improvement over LlamaGuard3 (8B) on English benchmarks while being 4.5x faster at inference with a significantly smaller model (0.5B). We achieve substantial advancements in multilingual safety tasks, particularly in addressing the imbalance for lower-resource languages in a collected real dataset. Ablation studies emphasize the critical role of synthetic data generation in bridging the imbalance in open-source data between English and other languages. These findings establish a scalable and efficient approach to synthetic data generation, paving the way for improved multilingual guardrail models to enhance LLM safety. Code, model, and data will be open-sourced at https://github.com/yihedeng9/DuoGuard. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05163v1-abstract-full').style.display = 'none'; document.getElementById('2502.05163v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 7 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">24 pages, 9 figures, 5 tables</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04411">arXiv:2502.04411</a> <span> [<a href="https://arxiv.org/pdf/2502.04411">pdf</a>, <a href="https://arxiv.org/format/2502.04411">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"> Mediator: Memory-efficient LLM Merging with Less Parameter Conflicts and Uncertainty Based Routing </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lai%2C+K">Kunfeng Lai</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+Z">Zhenheng Tang</a>, <a href="/search/cs?searchtype=author&query=Pan%2C+X">Xinglin Pan</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+P">Peijie Dong</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xiang Liu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+H">Haolan Chen</a>, <a href="/search/cs?searchtype=author&query=Shen%2C+L">Li Shen</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+X">Xiaowen Chu</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.04411v2-abstract-short" style="display: inline;"> Model merging aggregates Large Language Models (LLMs) finetuned on different tasks into a stronger one. However, parameter conflicts between models leads to performance degradation in averaging. While model routing addresses this issue by selecting individual models during inference, it imposes excessive storage and compute costs, and fails to leverage the common knowledge from different models. I… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04411v2-abstract-full').style.display = 'inline'; document.getElementById('2502.04411v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04411v2-abstract-full" style="display: none;"> Model merging aggregates Large Language Models (LLMs) finetuned on different tasks into a stronger one. However, parameter conflicts between models leads to performance degradation in averaging. While model routing addresses this issue by selecting individual models during inference, it imposes excessive storage and compute costs, and fails to leverage the common knowledge from different models. In this work, we observe that different layers exhibit varying levels of parameter conflicts. Building on this insight, we average layers with minimal parameter conflicts and use a novel task-level expert routing for layers with significant conflicts. To further reduce storage costs, inspired by task arithmetic sparsity, we decouple multiple fine-tuned experts into a dense expert and several sparse experts. Considering the out-of-distribution samples, we select and merge appropriate experts based on the task uncertainty of the input data. We conduct extensive experiments on both LLaMA and Qwen with varying parameter scales, and evaluate on real-world reasoning tasks. Results demonstrate that our method consistently achieves significant performance improvements while requiring less system cost compared to existing methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04411v2-abstract-full').style.display = 'none'; document.getElementById('2502.04411v2-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">v1</span> submitted 6 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. arXiv admin note: text overlap with arXiv:2405.09673 by other authors</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">MSC Class:</span> 68T50 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04176">arXiv:2502.04176</a> <span> [<a href="https://arxiv.org/pdf/2502.04176">pdf</a>, <a href="https://arxiv.org/format/2502.04176">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="Information Retrieval">cs.IR</span> </div> </div> <p class="title is-5 mathjax"> MRAMG-Bench: A BeyondText Benchmark for Multimodal Retrieval-Augmented Multimodal Generation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yu%2C+Q">Qinhan Yu</a>, <a href="/search/cs?searchtype=author&query=Xiao%2C+Z">Zhiyou Xiao</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Binghui Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhengren Wang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+C">Chong Chen</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+W">Wentao Zhang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.04176v1-abstract-short" style="display: inline;"> Recent advancements in Retrieval-Augmented Generation (RAG) have shown remarkable performance in enhancing response accuracy and relevance by integrating external knowledge into generative models. However, existing RAG methods primarily focus on providing text-only answers, even in multimodal retrieval-augmented generation scenarios. In this work, we introduce the Multimodal Retrieval-Augmented Mu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04176v1-abstract-full').style.display = 'inline'; document.getElementById('2502.04176v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04176v1-abstract-full" style="display: none;"> Recent advancements in Retrieval-Augmented Generation (RAG) have shown remarkable performance in enhancing response accuracy and relevance by integrating external knowledge into generative models. However, existing RAG methods primarily focus on providing text-only answers, even in multimodal retrieval-augmented generation scenarios. In this work, we introduce the Multimodal Retrieval-Augmented Multimodal Generation (MRAMG) task, which aims to generate answers that combine both text and images, fully leveraging the multimodal data within a corpus. Despite the importance of this task, there is a notable absence of a comprehensive benchmark to effectively evaluate MRAMG performance. To bridge this gap, we introduce the MRAMG-Bench, a carefully curated, human-annotated dataset comprising 4,346 documents, 14,190 images, and 4,800 QA pairs, sourced from three categories: Web Data, Academic Papers, and Lifestyle. The dataset incorporates diverse difficulty levels and complex multi-image scenarios, providing a robust foundation for evaluating multimodal generation tasks. To facilitate rigorous evaluation, our MRAMG-Bench incorporates a comprehensive suite of both statistical and LLM-based metrics, enabling a thorough analysis of the performance of popular generative models in the MRAMG task. Besides, we propose an efficient multimodal answer generation framework that leverages both LLMs and MLLMs to generate multimodal responses. Our datasets are available at: https://huggingface.co/MRAMG. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04176v1-abstract-full').style.display = 'none'; document.getElementById('2502.04176v1-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> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">11 pages</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04077">arXiv:2502.04077</a> <span> [<a href="https://arxiv.org/pdf/2502.04077">pdf</a>, <a href="https://arxiv.org/format/2502.04077">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> AttentionPredictor: Temporal Pattern Matters for Efficient LLM Inference </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yang%2C+Q">Qingyue Yang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+J">Jie Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xing Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhihai Wang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+C">Chen Chen</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+L">Lei Chen</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+X">Xianzhi Yu</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+W">Wulong Liu</a>, <a href="/search/cs?searchtype=author&query=Hao%2C+J">Jianye Hao</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+M">Mingxuan Yuan</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin 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="2502.04077v1-abstract-short" style="display: inline;"> With the development of large language models (LLMs), efficient inference through Key-Value (KV) cache compression has attracted considerable attention, especially for long-context generation. To compress the KV cache, recent methods identify critical KV tokens through heuristic ranking with attention scores. However, these methods often struggle to accurately determine critical tokens as they neg… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04077v1-abstract-full').style.display = 'inline'; document.getElementById('2502.04077v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04077v1-abstract-full" style="display: none;"> With the development of large language models (LLMs), efficient inference through Key-Value (KV) cache compression has attracted considerable attention, especially for long-context generation. To compress the KV cache, recent methods identify critical KV tokens through heuristic ranking with attention scores. However, these methods often struggle to accurately determine critical tokens as they neglect the \textit{temporal patterns} in attention scores, resulting in a noticeable degradation in LLM performance. To address this challenge, we propose AttentionPredictor, which is the first learning-based critical token identification approach. Specifically, AttentionPredictor learns a lightweight convolution model to capture spatiotemporal patterns and predict the next-token attention score. An appealing feature of AttentionPredictor is that it accurately predicts the attention score while consuming negligible memory. Moreover, we propose a cross-token critical cache prefetching framework that hides the token estimation time overhead to accelerate the decoding stage. By retaining most of the attention information, AttentionPredictor achieves 16$\times$ KV cache compression with comparable LLM performance, significantly outperforming the state-of-the-art. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04077v1-abstract-full').style.display = 'none'; document.getElementById('2502.04077v1-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/2502.03132">arXiv:2502.03132</a> <span> [<a href="https://arxiv.org/pdf/2502.03132">pdf</a>, <a href="https://arxiv.org/format/2502.03132">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Systems and Control">eess.SY</span> </div> </div> <p class="title is-5 mathjax"> SPARK: A Modular Benchmark for Humanoid Robot Safety </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sun%2C+Y">Yifan Sun</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+R">Rui Chen</a>, <a href="/search/cs?searchtype=author&query=Yun%2C+K+S">Kai S. Yun</a>, <a href="/search/cs?searchtype=author&query=Fang%2C+Y">Yikuan Fang</a>, <a href="/search/cs?searchtype=author&query=Jung%2C+S">Sebin Jung</a>, <a href="/search/cs?searchtype=author&query=Li%2C+F">Feihan Li</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bowei Li</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+W">Weiye Zhao</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+C">Changliu Liu</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.03132v1-abstract-short" style="display: inline;"> This paper introduces the Safe Protective and Assistive Robot Kit (SPARK), a comprehensive benchmark designed to ensure safety in humanoid autonomy and teleoperation. Humanoid robots pose significant safety risks due to their physical capabilities of interacting with complex environments. The physical structures of humanoid robots further add complexity to the design of general safety solutions. T… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03132v1-abstract-full').style.display = 'inline'; document.getElementById('2502.03132v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.03132v1-abstract-full" style="display: none;"> This paper introduces the Safe Protective and Assistive Robot Kit (SPARK), a comprehensive benchmark designed to ensure safety in humanoid autonomy and teleoperation. Humanoid robots pose significant safety risks due to their physical capabilities of interacting with complex environments. The physical structures of humanoid robots further add complexity to the design of general safety solutions. To facilitate the safe deployment of complex robot systems, SPARK can be used as a toolbox that comes with state-of-the-art safe control algorithms in a modular and composable robot control framework. Users can easily configure safety criteria and sensitivity levels to optimize the balance between safety and performance. To accelerate humanoid safety research and development, SPARK provides a simulation benchmark that compares safety approaches in a variety of environments, tasks, and robot models. Furthermore, SPARK allows quick deployment of synthesized safe controllers on real robots. For hardware deployment, SPARK supports Apple Vision Pro (AVP) or a Motion Capture System as external sensors, while also offering interfaces for seamless integration with alternative hardware setups. This paper demonstrates SPARK's capability with both simulation experiments and case studies with a Unitree G1 humanoid robot. Leveraging these advantages of SPARK, users and researchers can significantly improve the safety of their humanoid systems as well as accelerate relevant research. The open-source code is available at https://github.com/intelligent-control-lab/spark. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03132v1-abstract-full').style.display = 'none'; document.getElementById('2502.03132v1-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/2502.02441">arXiv:2502.02441</a> <span> [<a href="https://arxiv.org/pdf/2502.02441">pdf</a>, <a href="https://arxiv.org/format/2502.02441">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Multimedia">cs.MM</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"> LLMER: Crafting Interactive Extended Reality Worlds with JSON Data Generated by Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+J">Jiangong Chen</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+X">Xiaoyi Wu</a>, <a href="/search/cs?searchtype=author&query=Lan%2C+T">Tian Lan</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bin 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="2502.02441v1-abstract-short" style="display: inline;"> The integration of Large Language Models (LLMs) like GPT-4 with Extended Reality (XR) technologies offers the potential to build truly immersive XR environments that interact with human users through natural language, e.g., generating and animating 3D scenes from audio inputs. However, the complexity of XR environments makes it difficult to accurately extract relevant contextual data and scene/obj… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02441v1-abstract-full').style.display = 'inline'; document.getElementById('2502.02441v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.02441v1-abstract-full" style="display: none;"> The integration of Large Language Models (LLMs) like GPT-4 with Extended Reality (XR) technologies offers the potential to build truly immersive XR environments that interact with human users through natural language, e.g., generating and animating 3D scenes from audio inputs. However, the complexity of XR environments makes it difficult to accurately extract relevant contextual data and scene/object parameters from an overwhelming volume of XR artifacts. It leads to not only increased costs with pay-per-use models, but also elevated levels of generation errors. Moreover, existing approaches focusing on coding script generation are often prone to generation errors, resulting in flawed or invalid scripts, application crashes, and ultimately a degraded user experience. To overcome these challenges, we introduce LLMER, a novel framework that creates interactive XR worlds using JSON data generated by LLMs. Unlike prior approaches focusing on coding script generation, LLMER translates natural language inputs into JSON data, significantly reducing the likelihood of application crashes and processing latency. It employs a multi-stage strategy to supply only the essential contextual information adapted to the user's request and features multiple modules designed for various XR tasks. Our preliminary user study reveals the effectiveness of the proposed system, with over 80% reduction in consumed tokens and around 60% reduction in task completion time compared to state-of-the-art approaches. The analysis of users' feedback also illuminates a series of directions for further optimization. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02441v1-abstract-full').style.display = 'none'; document.getElementById('2502.02441v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 4 February, 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=Li%2C+B&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Li%2C+B&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