Search | arXiv e-print repository

<!DOCTYPE html> <html lang="en"> <head> <meta charset="utf-8"/> <meta name="viewport" content="width=device-width, initial-scale=1"/>  <link rel="apple-touch-icon" sizes="180x180" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/apple-touch-icon.png"> <link rel="icon" type="image/png" sizes="32x32" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon-32x32.png"> <link rel="icon" type="image/png" sizes="16x16" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon-16x16.png"> <link rel="manifest" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/site.webmanifest"> <link rel="mask-icon" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/safari-pinned-tab.svg" color="#b31b1b"> <link rel="shortcut icon" href="https://static.arxiv.org/static/base/1.0.0a5/images/icons/favicon.ico"> <meta name="msapplication-TileColor" content="#b31b1b"> <meta name="msapplication-config" content="images/icons/browserconfig.xml"> <meta name="theme-color" content="#b31b1b">  <title>Search | arXiv e-print repository</title> <script defer src="https://static.arxiv.org/static/base/1.0.0a5/fontawesome-free-5.11.2-web/js/all.js"></script> <link rel="stylesheet" href="https://static.arxiv.org/static/base/1.0.0a5/css/arxivstyle.css" /> <script type="text/x-mathjax-config"> MathJax.Hub.Config({ messageStyle: "none", extensions: ["tex2jax.js"], jax: ["input/TeX", "output/HTML-CSS"], tex2jax: { inlineMath: [ ['$','$'], ["\$","\$"] ], displayMath: [ ['$$','$$'], ["\\[","\\]"] ], processEscapes: true, ignoreClass: '.*', processClass: 'mathjax.*' }, TeX: { extensions: ["AMSmath.js", "AMSsymbols.js", "noErrors.js"], noErrors: { inlineDelimiters: ["$","$"], multiLine: false, style: { "font-size": "normal", "border": "" } } }, "HTML-CSS": { availableFonts: ["TeX"] } }); </script> <script src='//static.arxiv.org/MathJax-2.7.3/MathJax.js'></script> <script src="https://static.arxiv.org/static/base/1.0.0a5/js/notification.js"></script> <link rel="stylesheet" href="https://static.arxiv.org/static/search/0.5.6/css/bulma-tooltip.min.css" /> <link rel="stylesheet" href="https://static.arxiv.org/static/search/0.5.6/css/search.css" /> <script src="https://code.jquery.com/jquery-3.2.1.slim.min.js" integrity="sha256-k2WSCIexGzOj3Euiig+TlR8gA0EmPjuc79OEeY5L45g=" crossorigin="anonymous"></script> <script src="https://static.arxiv.org/static/search/0.5.6/js/fieldset.js"></script> <style> radio#cf-customfield_11400 { display: none; } </style> </head> <body> <header><a href="#main-container" class="is-sr-only">Skip to main content</a>  <div class="attribution level is-marginless" role="banner"> <div class="level-left"> <a class="level-item" href="https://cornell.edu/"><img src="https://static.arxiv.org/static/base/1.0.0a5/images/cornell-reduced-white-SMALL.svg" alt="Cornell University" width="200" aria-label="logo" /></a> </div> <div class="level-right is-marginless"><p class="sponsors level-item is-marginless"><span id="support-ack-url">We gratefully acknowledge support from<br /> the Simons Foundation, <a href="https://info.arxiv.org/about/ourmembers.html">member institutions</a>, and all contributors. <a href="https://info.arxiv.org/about/donate.html">Donate</a></span></p></div> </div>  <div class="identity level is-marginless"> <div class="level-left"> <div class="level-item"> <a class="arxiv" href="https://arxiv.org/" aria-label="arxiv-logo"> <img src="https://static.arxiv.org/static/base/1.0.0a5/images/arxiv-logo-one-color-white.svg" aria-label="logo" alt="arxiv logo" width="85" style="width:85px;"/> </a> </div> </div> <div class="search-block level-right"> <form class="level-item mini-search" method="GET" action="https://arxiv.org/search"> <div class="field has-addons"> <div class="control"> <input class="input is-small" type="text" name="query" placeholder="Search..." aria-label="Search term or terms" /> <p class="help"><a href="https://info.arxiv.org/help">Help</a> | <a href="https://arxiv.org/search/advanced">Advanced Search</a></p> </div> <div class="control"> <div class="select is-small"> <select name="searchtype" aria-label="Field to search"> <option value="all" selected="selected">All fields</option> <option value="title">Title</option> <option value="author">Author</option> <option value="abstract">Abstract</option> <option value="comments">Comments</option> <option value="journal_ref">Journal reference</option> <option value="acm_class">ACM classification</option> <option value="msc_class">MSC classification</option> <option value="report_num">Report number</option> <option value="paper_id">arXiv identifier</option> <option value="doi">DOI</option> <option value="orcid">ORCID</option> <option value="author_id">arXiv author ID</option> <option value="help">Help pages</option> <option value="full_text">Full text</option> </select> </div> </div> <input type="hidden" name="source" value="header"> <button class="button is-small is-cul-darker">Search</button> </div> </form> </div> </div>  <div class="container"> <div class="user-tools is-size-7 has-text-right has-text-weight-bold" role="navigation" aria-label="User menu"> <a href="https://arxiv.org/login">Login</a> </div> </div> </header> <main class="container" id="main-container"> <div class="level is-marginless"> <div class="level-left"> <h1 class="title is-clearfix"> Showing 1–50 of 1,948 results for author: <span class="mathjax">Wang, R</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=Wang%2C+R">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="Wang, R"> </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=Wang%2C+R&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="Wang, R"> <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=Wang%2C+R&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&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/2502.18273">arXiv:2502.18273</a> <span> [<a href="https://arxiv.org/pdf/2502.18273">pdf</a>, <a href="https://arxiv.org/format/2502.18273">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"> Beyond In-Distribution Success: Scaling Curves of CoT Granularity for Language Model Generalization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ru Wang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+W">Wei Huang</a>, <a href="/search/cs?searchtype=author&query=Song%2C+S">Selena Song</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Haoyu Zhang</a>, <a href="/search/cs?searchtype=author&query=Iwasawa%2C+Y">Yusuke Iwasawa</a>, <a href="/search/cs?searchtype=author&query=Matsuo%2C+Y">Yutaka Matsuo</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+J">Jiaxian Guo</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.18273v1-abstract-short" style="display: inline;"> Generalization to novel compound tasks under distribution shift is important for deploying transformer-based language models (LMs). This work investigates Chain-of-Thought (CoT) reasoning as a means to enhance OOD generalization. Through controlled experiments across several compound tasks, we reveal three key insights: (1) While QA-trained models achieve near-perfect in-distribution accuracy, the… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18273v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18273v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18273v1-abstract-full" style="display: none;"> Generalization to novel compound tasks under distribution shift is important for deploying transformer-based language models (LMs). This work investigates Chain-of-Thought (CoT) reasoning as a means to enhance OOD generalization. Through controlled experiments across several compound tasks, we reveal three key insights: (1) While QA-trained models achieve near-perfect in-distribution accuracy, their OOD performance degrades catastrophically, even with 10000k+ training examples; (2) the granularity of CoT data strongly correlates with generalization performance; finer-grained CoT data leads to better generalization; (3) CoT exhibits remarkable sample efficiency, matching QA performance with much less (even 80%) data. Theoretically, we demonstrate that compound tasks inherently permit shortcuts in Q-A data that misalign with true reasoning principles, while CoT forces internalization of valid dependency structures, and thus can achieve better generalization. Further, we show that transformer positional embeddings can amplify generalization by emphasizing subtask condition recurrence in long CoT sequences. Our combined theoretical and empirical analysis provides compelling evidence for CoT reasoning as a crucial training paradigm for enabling LM generalization under real-world distributional shifts for compound tasks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18273v1-abstract-full').style.display = 'none'; document.getElementById('2502.18273v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.18228">arXiv:2502.18228</a> <span> [<a href="https://arxiv.org/pdf/2502.18228">pdf</a>, <a href="https://arxiv.org/format/2502.18228">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"> Debt Collection Negotiations with Large Language Models: An Evaluation System and Optimizing Decision Making with Multi-Agent </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xiaofeng Wang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhixin Zhang</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+J">Jinguang Zheng</a>, <a href="/search/cs?searchtype=author&query=Ai%2C+Y">Yiming Ai</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui 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="2502.18228v1-abstract-short" style="display: inline;"> Debt collection negotiations (DCN) are vital for managing non-performing loans (NPLs) and reducing creditor losses. Traditional methods are labor-intensive, while large language models (LLMs) offer promising automation potential. However, prior systems lacked dynamic negotiation and real-time decision-making capabilities. This paper explores LLMs in automating DCN and proposes a novel evaluation f… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18228v1-abstract-full').style.display = 'inline'; document.getElementById('2502.18228v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.18228v1-abstract-full" style="display: none;"> Debt collection negotiations (DCN) are vital for managing non-performing loans (NPLs) and reducing creditor losses. Traditional methods are labor-intensive, while large language models (LLMs) offer promising automation potential. However, prior systems lacked dynamic negotiation and real-time decision-making capabilities. This paper explores LLMs in automating DCN and proposes a novel evaluation framework with 13 metrics across 4 aspects. Our experiments reveal that LLMs tend to over-concede compared to human negotiators. To address this, we propose the Multi-Agent Debt Negotiation (MADeN) framework, incorporating planning and judging modules to improve decision rationality. We also apply post-training techniques, including DPO with rejection sampling, to optimize performance. Our studies provide valuable insights for practitioners and researchers seeking to enhance efficiency and outcomes in this domain. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.18228v1-abstract-full').style.display = 'none'; document.getElementById('2502.18228v1-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">21 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/2502.17874">arXiv:2502.17874</a> <span> [<a href="https://arxiv.org/pdf/2502.17874">pdf</a>, <a href="https://arxiv.org/format/2502.17874">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="Quantitative Methods">q-bio.QM</span> </div> </div> <p class="title is-5 mathjax"> Neural Graph Matching Improves Retrieval Augmented Generation in Molecular Machine Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzhong Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui-Xi Wang</a>, <a href="/search/cs?searchtype=author&query=Manjrekar%2C+M">Mrunali Manjrekar</a>, <a href="/search/cs?searchtype=author&query=Coley%2C+C+W">Connor W. Coley</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="2502.17874v1-abstract-short" style="display: inline;"> Molecular machine learning has gained popularity with the advancements of geometric deep learning. In parallel, retrieval-augmented generation has become a principled approach commonly used with language models. However, the optimal integration of retrieval augmentation into molecular machine learning remains unclear. Graph neural networks stand to benefit from clever matching to understand the st… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17874v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17874v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17874v1-abstract-full" style="display: none;"> Molecular machine learning has gained popularity with the advancements of geometric deep learning. In parallel, retrieval-augmented generation has become a principled approach commonly used with language models. However, the optimal integration of retrieval augmentation into molecular machine learning remains unclear. Graph neural networks stand to benefit from clever matching to understand the structural alignment of retrieved molecules to a query molecule. Neural graph matching offers a compelling solution by explicitly modeling node and edge affinities between two structural graphs while employing a noise-robust, end-to-end neural network to learn affinity metrics. We apply this approach to mass spectrum simulation and introduce MARASON, a novel model that incorporates neural graph matching to enhance a fragmentation-based neural network. Experimental results highlight the effectiveness of our design, with MARASON achieving 28% top-1 accuracy, a substantial improvement over the non-retrieval state-of-the-art accuracy of 19%. Moreover, MARASON outperforms both naive retrieval-augmented generation methods and traditional graph matching approaches. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17874v1-abstract-full').style.display = 'none'; document.getElementById('2502.17874v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.17529">arXiv:2502.17529</a> <span> [<a href="https://arxiv.org/pdf/2502.17529">pdf</a>, <a href="https://arxiv.org/format/2502.17529">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Multiagent Systems">cs.MA</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> </div> </div> <p class="title is-5 mathjax"> ConvoyLLM: Dynamic Multi-Lane Convoy Control Using LLMs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Lu%2C+L">Liping Lu</a>, <a href="/search/cs?searchtype=author&query=He%2C+Z">Zhican He</a>, <a href="/search/cs?searchtype=author&query=Chu%2C+D">Duanfeng Chu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rukang Wang</a>, <a href="/search/cs?searchtype=author&query=Peng%2C+S">Saiqian Peng</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+P">Pan 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="2502.17529v1-abstract-short" style="display: inline;"> This paper proposes a novel method for multi-lane convoy formation control that uses large language models (LLMs) to tackle coordination challenges in dynamic highway environments. Each connected and autonomous vehicle in the convoy uses a knowledge-driven approach to make real-time adaptive decisions based on various scenarios. Our method enables vehicles to dynamically perform tasks, including o… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17529v1-abstract-full').style.display = 'inline'; document.getElementById('2502.17529v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.17529v1-abstract-full" style="display: none;"> This paper proposes a novel method for multi-lane convoy formation control that uses large language models (LLMs) to tackle coordination challenges in dynamic highway environments. Each connected and autonomous vehicle in the convoy uses a knowledge-driven approach to make real-time adaptive decisions based on various scenarios. Our method enables vehicles to dynamically perform tasks, including obstacle avoidance, convoy joining/leaving, and escort formation switching, all while maintaining the overall convoy structure. We design a Interlaced formation control strategy based on locally dynamic distributed graphs, ensuring the convoy remains stable and flexible. We conduct extensive experiments in the SUMO simulation platform across multiple traffic scenarios, and the results demonstrate that the proposed method is effective, robust, and adaptable to dynamic environments. The code is available at: https://github.com/chuduanfeng/ConvoyLLM. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.17529v1-abstract-full').style.display = 'none'; document.getElementById('2502.17529v1-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/2502.15806">arXiv:2502.15806</a> <span> [<a href="https://arxiv.org/pdf/2502.15806">pdf</a>, <a href="https://arxiv.org/format/2502.15806">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> A Mousetrap: Fooling Large Reasoning Models for Jailbreak with Chain of Iterative Chaos </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Yao%2C+Y">Yang Yao</a>, <a href="/search/cs?searchtype=author&query=Tong%2C+X">Xuan Tong</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruofan Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yixu Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+L">Lujundong Li</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+L">Liang Liu</a>, <a href="/search/cs?searchtype=author&query=Teng%2C+Y">Yan Teng</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yingchun 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="2502.15806v1-abstract-short" style="display: inline;"> Large Reasoning Models (LRMs) have significantly advanced beyond traditional Large Language Models (LLMs) with their exceptional logical reasoning capabilities, yet these improvements introduce heightened safety risks. When subjected to jailbreak attacks, their ability to generate more targeted and organized content can lead to greater harm. Although some studies claim that reasoning enables safer… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15806v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15806v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15806v1-abstract-full" style="display: none;"> Large Reasoning Models (LRMs) have significantly advanced beyond traditional Large Language Models (LLMs) with their exceptional logical reasoning capabilities, yet these improvements introduce heightened safety risks. When subjected to jailbreak attacks, their ability to generate more targeted and organized content can lead to greater harm. Although some studies claim that reasoning enables safer LRMs against existing LLM attacks, they overlook the inherent flaws within the reasoning process itself. To address this gap, we propose the first jailbreak attack targeting LRMs, exploiting their unique vulnerabilities stemming from the advanced reasoning capabilities. Specifically, we introduce a Chaos Machine, a novel component to transform attack prompts with diverse one-to-one mappings. The chaos mappings iteratively generated by the machine are embedded into the reasoning chain, which strengthens the variability and complexity and also promotes a more robust attack. Based on this, we construct the Mousetrap framework, which makes attacks projected into nonlinear-like low sample spaces with mismatched generalization enhanced. Also, due to the more competing objectives, LRMs gradually maintain the inertia of unpredictable iterative reasoning and fall into our trap. Success rates of the Mousetrap attacking o1-mini, claude-sonnet and gemini-thinking are as high as 96%, 86% and 98% respectively on our toxic dataset Trotter. On benchmarks such as AdvBench, StrongREJECT, and HarmBench, attacking claude-sonnet, well-known for its safety, Mousetrap can astonishingly achieve success rates of 87.5%, 86.58% and 93.13% respectively. Attention: This paper contains inappropriate, offensive and harmful content. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15806v1-abstract-full').style.display = 'none'; document.getElementById('2502.15806v1-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/2502.15706">arXiv:2502.15706</a> <span> [<a href="https://arxiv.org/pdf/2502.15706">pdf</a>, <a href="https://arxiv.org/format/2502.15706">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Networking and Internet Architecture">cs.NI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Signal Processing">eess.SP</span> </div> </div> <p class="title is-5 mathjax"> Multi-Failure Localization in High-Degree ROADM-based Optical Networks using Rules-Informed Neural Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruikun Wang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Q">Qiaolun Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jiawei Zhang</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+Z">Zhiqun Gu</a>, <a href="/search/cs?searchtype=author&query=Ibrahimi%2C+M">Memedhe Ibrahimi</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+H">Hao Yu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+B">Bojun Zhang</a>, <a href="/search/cs?searchtype=author&query=Musumeci%2C+F">Francesco Musumeci</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+Y">Yuefeng Ji</a>, <a href="/search/cs?searchtype=author&query=Tornatore%2C+M">Massimo Tornatore</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="2502.15706v1-abstract-short" style="display: inline;"> To accommodate ever-growing traffic, network operators are actively deploying high-degree reconfigurable optical add/drop multiplexers (ROADMs) to build large-capacity optical networks. High-degree ROADM-based optical networks have multiple parallel fibers between ROADM nodes, requiring the adoption of ROADM nodes with a large number of inter-/intra-node components. However, this large number of i… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15706v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15706v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15706v1-abstract-full" style="display: none;"> To accommodate ever-growing traffic, network operators are actively deploying high-degree reconfigurable optical add/drop multiplexers (ROADMs) to build large-capacity optical networks. High-degree ROADM-based optical networks have multiple parallel fibers between ROADM nodes, requiring the adoption of ROADM nodes with a large number of inter-/intra-node components. However, this large number of inter-/intra-node optical components in high-degree ROADM networks increases the likelihood of multiple failures simultaneously, and calls for novel methods for accurate localization of multiple failed components. To the best of our knowledge, this is the first study investigating the problem of multi-failure localization for high-degree ROADM-based optical networks. To solve this problem, we first provide a description of the failures affecting both inter-/intra-node components, and we consider different deployments of optical power monitors (OPMs) to obtain information (i.e., optical power) to be used for automated multi-failure localization. Then, as our main and original contribution, we propose a novel method based on a rules-informed neural network (RINN) for multi-failure localization, which incorporates the benefits of both rules-based reasoning and artificial neural networks (ANN). Through extensive simulations and experimental demonstrations, we show that our proposed RINN algorithm can achieve up to around 20 higher localization accuracy compared to baseline algorithms, incurring only around 4.14 ms of average inference time. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15706v1-abstract-full').style.display = 'none'; document.getElementById('2502.15706v1-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 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">This is the author's version of the work. This work was accepted by IEEE Journal on Selected Areas in Communications</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Journal ref:</span> IEEE Journal on Selected Areas in Communications, 2025 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.15302">arXiv:2502.15302</a> <span> [<a href="https://arxiv.org/pdf/2502.15302">pdf</a>, <a href="https://arxiv.org/format/2502.15302">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"> A Novel Riemannian Sparse Representation Learning Network for Polarimetric SAR Image Classification </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shi%2C+J">Junfei Shi</a>, <a href="/search/cs?searchtype=author&query=Nie%2C+M">Mengmeng Nie</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+W">Weisi Lin</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+H">Haiyan Jin</a>, <a href="/search/cs?searchtype=author&query=Li%2C+J">Junhuai Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui 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="2502.15302v1-abstract-short" style="display: inline;"> Deep learning is an effective end-to-end method for Polarimetric Synthetic Aperture Radar(PolSAR) image classification, but it lacks the guidance of related mathematical principle and is essentially a black-box model. In addition, existing deep models learn features in Euclidean space, where PolSAR complex matrix is commonly converted into a complex-valued vector as the network input, distorting m… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15302v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15302v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15302v1-abstract-full" style="display: none;"> Deep learning is an effective end-to-end method for Polarimetric Synthetic Aperture Radar(PolSAR) image classification, but it lacks the guidance of related mathematical principle and is essentially a black-box model. In addition, existing deep models learn features in Euclidean space, where PolSAR complex matrix is commonly converted into a complex-valued vector as the network input, distorting matrix structure and channel relationship. However, the complex covariance matrix is Hermitian positive definite (HPD), and resides on a Riemannian manifold instead of a Euclidean one. Existing methods cannot measure the geometric distance of HPD matrices and easily cause some misclassifications due to inappropriate Euclidean measures. To address these issues, we propose a novel Riemannian Sparse Representation Learning Network (SRSR CNN) for PolSAR images. Firstly, a superpixel-based Riemannian Sparse Representation (SRSR) model is designed to learn the sparse features with Riemannian metric. Then, the optimization procedure of the SRSR model is inferred and further unfolded into an SRSRnet, which can automatically learn the sparse coefficients and dictionary atoms. Furthermore, to learn contextual high-level features, a CNN-enhanced module is added to improve classification performance. The proposed network is a Sparse Representation (SR) guided deep learning model, which can directly utilize the covariance matrix as the network input, and utilize Riemannian metric to learn geometric structure and sparse features of complex matrices in Riemannian space. Experiments on three real PolSAR datasets demonstrate that the proposed method surpasses state-of-the-art techniques in ensuring accurate edge details and correct region homogeneity for classification. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15302v1-abstract-full').style.display = 'none'; document.getElementById('2502.15302v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 21 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">13 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/2502.15285">arXiv:2502.15285</a> <span> [<a href="https://arxiv.org/pdf/2502.15285">pdf</a>, <a href="https://arxiv.org/format/2502.15285">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Sound">cs.SD</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey 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="Networking and Internet Architecture">cs.NI</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"> Offload Rethinking by Cloud Assistance for Efficient Environmental Sound Recognition on LPWANs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+L">Le Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+Q">Quanling Zhao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Run Wang</a>, <a href="/search/cs?searchtype=author&query=Bian%2C+S">Shirley Bian</a>, <a href="/search/cs?searchtype=author&query=Gungor%2C+O">Onat Gungor</a>, <a href="/search/cs?searchtype=author&query=Ponzina%2C+F">Flavio Ponzina</a>, <a href="/search/cs?searchtype=author&query=Rosing%2C+T">Tajana Rosing</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="2502.15285v1-abstract-short" style="display: inline;"> Learning-based environmental sound recognition has emerged as a crucial method for ultra-low-power environmental monitoring in biological research and city-scale sensing systems. These systems usually operate under limited resources and are often powered by harvested energy in remote areas. Recent efforts in on-device sound recognition suffer from low accuracy due to resource constraints, whereas… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15285v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15285v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15285v1-abstract-full" style="display: none;"> Learning-based environmental sound recognition has emerged as a crucial method for ultra-low-power environmental monitoring in biological research and city-scale sensing systems. These systems usually operate under limited resources and are often powered by harvested energy in remote areas. Recent efforts in on-device sound recognition suffer from low accuracy due to resource constraints, whereas cloud offloading strategies are hindered by high communication costs. In this work, we introduce ORCA, a novel resource-efficient cloud-assisted environmental sound recognition system on batteryless devices operating over the Low-Power Wide-Area Networks (LPWANs), targeting wide-area audio sensing applications. We propose a cloud assistance strategy that remedies the low accuracy of on-device inference while minimizing the communication costs for cloud offloading. By leveraging a self-attention-based cloud sub-spectral feature selection method to facilitate efficient on-device inference, ORCA resolves three key challenges for resource-constrained cloud offloading over LPWANs: 1) high communication costs and low data rates, 2) dynamic wireless channel conditions, and 3) unreliable offloading. We implement ORCA on an energy-harvesting batteryless microcontroller and evaluate it in a real world urban sound testbed. Our results show that ORCA outperforms state-of-the-art methods by up to $80 \times$ in energy savings and $220 \times$ in latency reduction while maintaining comparable accuracy. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15285v1-abstract-full').style.display = 'none'; document.getElementById('2502.15285v1-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/2502.15260">arXiv:2502.15260</a> <span> [<a href="https://arxiv.org/pdf/2502.15260">pdf</a>, <a href="https://arxiv.org/format/2502.15260">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"> LightMamba: Efficient Mamba Acceleration on FPGA with Quantization and Hardware Co-design </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wei%2C+R">Renjie Wei</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+S">Songqiang Xu</a>, <a href="/search/cs?searchtype=author&query=Zhong%2C+L">Linfeng Zhong</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Z">Zebin Yang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+Q">Qingyu Guo</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yuan Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runsheng Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+M">Meng 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="2502.15260v1-abstract-short" style="display: inline;"> State space models (SSMs) like Mamba have recently attracted much attention. Compared to Transformer-based large language models (LLMs), Mamba achieves linear computation complexity with the sequence length and demonstrates superior performance. However, Mamba is hard to accelerate due to the scattered activation outliers and the complex computation dependency, rendering existing LLM accelerators… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15260v1-abstract-full').style.display = 'inline'; document.getElementById('2502.15260v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.15260v1-abstract-full" style="display: none;"> State space models (SSMs) like Mamba have recently attracted much attention. Compared to Transformer-based large language models (LLMs), Mamba achieves linear computation complexity with the sequence length and demonstrates superior performance. However, Mamba is hard to accelerate due to the scattered activation outliers and the complex computation dependency, rendering existing LLM accelerators inefficient. In this paper, we propose LightMamba that co-designs the quantization algorithm and FPGA accelerator architecture for efficient Mamba inference. We first propose an FPGA-friendly post-training quantization algorithm that features rotation-assisted quantization and power-of-two SSM quantization to reduce the majority of computation to 4-bit. We further design an FPGA accelerator that partially unrolls the Mamba computation to balance the efficiency and hardware costs. Through computation reordering as well as fine-grained tiling and fusion, the hardware utilization and memory efficiency of the accelerator get drastically improved. We implement LightMamba on Xilinx Versal VCK190 FPGA and achieve 4.65x to 6.06x higher energy efficiency over the GPU baseline. When evaluated on Alveo U280 FPGA, LightMamba reaches 93 tokens/s, which is 1.43x that of the GPU baseline. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.15260v1-abstract-full').style.display = 'none'; document.getElementById('2502.15260v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 21 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted by DATE 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/2502.14894">arXiv:2502.14894</a> <span> [<a href="https://arxiv.org/pdf/2502.14894">pdf</a>, <a href="https://arxiv.org/format/2502.14894">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="Computers and Society">cs.CY</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"> FOCUS on Contamination: A Geospatial Deep Learning Framework with a Noise-Aware Loss for Surface Water PFAS Prediction </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Khan%2C+J">Jowaria Khan</a>, <a href="/search/cs?searchtype=author&query=Friedman%2C+A">Alexa Friedman</a>, <a href="/search/cs?searchtype=author&query=Evans%2C+S">Sydney Evans</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzi Wang</a>, <a href="/search/cs?searchtype=author&query=Beins%2C+K">Kaley Beins</a>, <a href="/search/cs?searchtype=author&query=Andrews%2C+D">David Andrews</a>, <a href="/search/cs?searchtype=author&query=Bondi-Kelly%2C+E">Elizabeth Bondi-Kelly</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="2502.14894v1-abstract-short" style="display: inline;"> Per and polyfluoroalkyl substances (PFAS), chemicals found in products like non-stick cookware, are unfortunately persistent environmental pollutants with severe health risks. Accurately mapping PFAS contamination is crucial for guiding targeted remediation efforts and protecting public and environmental health, yet detection across large regions remains challenging due to the cost of testing and… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14894v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14894v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14894v1-abstract-full" style="display: none;"> Per and polyfluoroalkyl substances (PFAS), chemicals found in products like non-stick cookware, are unfortunately persistent environmental pollutants with severe health risks. Accurately mapping PFAS contamination is crucial for guiding targeted remediation efforts and protecting public and environmental health, yet detection across large regions remains challenging due to the cost of testing and the difficulty of simulating their spread. In this work, we introduce FOCUS, a geospatial deep learning framework with a label noise-aware loss function, to predict PFAS contamination in surface water over large regions. By integrating hydrological flow data, land cover information, and proximity to known PFAS sources, our approach leverages both spatial and environmental context to improve prediction accuracy. We evaluate the performance of our approach through extensive ablation studies and comparative analyses against baselines like sparse segmentation, as well as existing scientific methods, including Kriging and pollutant transport simulations. Results highlight our framework's potential for scalable PFAS monitoring. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14894v1-abstract-full').style.display = 'none'; document.getElementById('2502.14894v1-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">ACM Class:</span> I.2.1; I.2.10; I.4.6; I.4.9; I.4.10; J.2 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.14617">arXiv:2502.14617</a> <span> [<a href="https://arxiv.org/pdf/2502.14617">pdf</a>, <a href="https://arxiv.org/format/2502.14617">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"> Serving Models, Fast and Slow:Optimizing Heterogeneous LLM Inferencing Workloads at Scale </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Jaiswal%2C+S">Shashwat Jaiswal</a>, <a href="/search/cs?searchtype=author&query=Jain%2C+K">Kunal Jain</a>, <a href="/search/cs?searchtype=author&query=Simmhan%2C+Y">Yogesh Simmhan</a>, <a href="/search/cs?searchtype=author&query=Parayil%2C+A">Anjaly Parayil</a>, <a href="/search/cs?searchtype=author&query=Mallick%2C+A">Ankur Mallick</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rujia Wang</a>, <a href="/search/cs?searchtype=author&query=Amant%2C+R+S">Renee St. Amant</a>, <a href="/search/cs?searchtype=author&query=Bansal%2C+C">Chetan Bansal</a>, <a href="/search/cs?searchtype=author&query=R%C3%BChle%2C+V">Victor R眉hle</a>, <a href="/search/cs?searchtype=author&query=Kulkarni%2C+A">Anoop Kulkarni</a>, <a href="/search/cs?searchtype=author&query=Kofsky%2C+S">Steve Kofsky</a>, <a href="/search/cs?searchtype=author&query=Rajmohan%2C+S">Saravan Rajmohan</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="2502.14617v1-abstract-short" style="display: inline;"> Large Language Model (LLM) inference workloads handled by global cloud providers can include both latency-sensitive and insensitive tasks, creating a diverse range of Service Level Agreement (SLA) requirements. Managing these mixed workloads is challenging due to the complexity of the inference stack, which includes multiple LLMs, hardware configurations, and geographic distributions. Current opti… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14617v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14617v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14617v1-abstract-full" style="display: none;"> Large Language Model (LLM) inference workloads handled by global cloud providers can include both latency-sensitive and insensitive tasks, creating a diverse range of Service Level Agreement (SLA) requirements. Managing these mixed workloads is challenging due to the complexity of the inference stack, which includes multiple LLMs, hardware configurations, and geographic distributions. Current optimization strategies often silo these tasks to ensure that SLAs are met for latency-sensitive tasks, but this leads to significant under-utilization of expensive GPU resources despite the availability of spot and on-demand Virtual Machine (VM) provisioning. We propose SAGESERVE, a comprehensive LLM serving framework that employs adaptive control knobs at varying time scales, ensuring SLA compliance while maximizing the utilization of valuable GPU resources. Short-term optimizations include efficient request routing to data center regions, while long-term strategies involve scaling GPU VMs out/in and redeploying models to existing VMs to align with traffic patterns. These strategies are formulated as an optimization problem for resource allocation and solved using Integer Linear Programming (ILP). We perform empirical and simulation studies based on production workload traces with over 8M requests using four open-source models deployed across three regions. SAGESERVE achieves up to 25% savings in GPU-hours while maintaining tail latency and satisfying all SLOs, and it reduces the scaling overhead compared to baselines by up to 80%, confirming the effectiveness of our proposal. In terms of dollar cost, this can save cloud providers up to $2M over the course of a month. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14617v1-abstract-full').style.display = 'none'; document.getElementById('2502.14617v1-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, 17 figures, 2 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/2502.14583">arXiv:2502.14583</a> <span> [<a href="https://arxiv.org/pdf/2502.14583">pdf</a>, <a href="https://arxiv.org/format/2502.14583">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> A Theory for Conditional Generative Modeling on Multiple Data Sources </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rongzhen Wang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Y">Yan Zhang</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+C">Chenyu Zheng</a>, <a href="/search/cs?searchtype=author&query=Li%2C+C">Chongxuan Li</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+G">Guoqiang 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="2502.14583v1-abstract-short" style="display: inline;"> The success of large generative models has driven a paradigm shift, leveraging massive multi-source data to enhance model capabilities. However, the interaction among these sources remains theoretically underexplored. This paper takes the first step toward a rigorous analysis of multi-source training in conditional generative modeling, where each condition represents a distinct data source. Specif… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14583v1-abstract-full').style.display = 'inline'; document.getElementById('2502.14583v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.14583v1-abstract-full" style="display: none;"> The success of large generative models has driven a paradigm shift, leveraging massive multi-source data to enhance model capabilities. However, the interaction among these sources remains theoretically underexplored. This paper takes the first step toward a rigorous analysis of multi-source training in conditional generative modeling, where each condition represents a distinct data source. Specifically, we establish a general distribution estimation error bound in average total variation distance for conditional maximum likelihood estimation based on the bracketing number. Our result shows that when source distributions share certain similarities and the model is expressive enough, multi-source training guarantees a sharper bound than single-source training. We further instantiate the general theory on conditional Gaussian estimation and deep generative models including autoregressive and flexible energy-based models, by characterizing their bracketing numbers. The results highlight that the number of sources and similarity among source distributions improve the advantage of multi-source training. Simulations and real-world experiments validate our theory. Code is available at: \url{https://github.com/ML-GSAI/Multi-Source-GM}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.14583v1-abstract-full').style.display = 'none'; document.getElementById('2502.14583v1-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">35 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/2502.13753">arXiv:2502.13753</a> <span> [<a href="https://arxiv.org/pdf/2502.13753">pdf</a>, <a href="https://arxiv.org/format/2502.13753">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"> SCALAR: Scientific Citation-based Live Assessment of Long-context Academic Reasoning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Renxi Wang</a>, <a href="/search/cs?searchtype=author&query=Mu%2C+H">Honglin Mu</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+L">Liqun Ma</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+L">Lizhi Lin</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+Y">Yunlong Feng</a>, <a href="/search/cs?searchtype=author&query=Baldwin%2C+T">Timothy Baldwin</a>, <a href="/search/cs?searchtype=author&query=Han%2C+X">Xudong Han</a>, <a href="/search/cs?searchtype=author&query=Li%2C+H">Haonan 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="2502.13753v1-abstract-short" style="display: inline;"> Evaluating large language models' (LLMs) long-context understanding capabilities remains challenging. We present SCALAR (Scientific Citation-based Live Assessment of Long-context Academic Reasoning), a novel benchmark that leverages academic papers and their citation networks. SCALAR features automatic generation of high-quality ground truth labels without human annotation, controllable difficulty… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13753v1-abstract-full').style.display = 'inline'; document.getElementById('2502.13753v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.13753v1-abstract-full" style="display: none;"> Evaluating large language models' (LLMs) long-context understanding capabilities remains challenging. We present SCALAR (Scientific Citation-based Live Assessment of Long-context Academic Reasoning), a novel benchmark that leverages academic papers and their citation networks. SCALAR features automatic generation of high-quality ground truth labels without human annotation, controllable difficulty levels, and a dynamic updating mechanism that prevents data contamination. Using ICLR 2025 papers, we evaluate 8 state-of-the-art LLMs, revealing key insights about their capabilities and limitations in processing long scientific documents across different context lengths and reasoning types. Our benchmark provides a reliable and sustainable way to track progress in long-context understanding as LLM capabilities evolve. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.13753v1-abstract-full').style.display = 'none'; document.getElementById('2502.13753v1-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/2502.12202">arXiv:2502.12202</a> <span> [<a href="https://arxiv.org/pdf/2502.12202">pdf</a>, <a href="https://arxiv.org/format/2502.12202">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> BoT: Breaking Long Thought Processes of o1-like Large Language Models through Backdoor Attack </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhu%2C+Z">Zihao Zhu</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+H">Hongbao Zhang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+M">Mingda Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruotong Wang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+G">Guanzong Wu</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+K">Ke Xu</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+B">Baoyuan 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="2502.12202v1-abstract-short" style="display: inline;"> Longer thought, better performance: large language models with deep reasoning capabilities, particularly o1-like models, have demonstrated remarkable performance by generating extensive thought processes during inference. This trade-off reveals a potential vulnerability: adversaries could compromise model performance by forcing immediate responses without thought processes. To this end, in this pa… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12202v1-abstract-full').style.display = 'inline'; document.getElementById('2502.12202v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.12202v1-abstract-full" style="display: none;"> Longer thought, better performance: large language models with deep reasoning capabilities, particularly o1-like models, have demonstrated remarkable performance by generating extensive thought processes during inference. This trade-off reveals a potential vulnerability: adversaries could compromise model performance by forcing immediate responses without thought processes. To this end, in this paper, we introduce a novel attack scenario targeting the long thought processes of o1-like models and propose BoT (Break CoT), which can selectively break intrinsic reasoning mechanisms through backdoor attacks. BoT constructs poisoned datasets with designed triggers and injects backdoor by either supervised fine-tuning or direct preference optimization. When triggered, the model directly generates answers without thought processes, while maintaining normal reasoning capabilities for clean inputs. Extensive experiments on open-source o1-like models, including recent DeepSeek-R1, demonstrate that BoT nearly achieves high attack success rates while maintaining clean accuracy, highlighting the critical safety risk in current models. Furthermore, the relationship between task difficulty and helpfulness reveals a potential application for good, enabling users to customize model behavior based on task complexity. Code is available at \href{https://github.com/zihao-ai/BoT}{https://github.com/zihao-ai/BoT}. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.12202v1-abstract-full').style.display = 'none'; document.getElementById('2502.12202v1-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/2502.11740">arXiv:2502.11740</a> <span> [<a href="https://arxiv.org/pdf/2502.11740">pdf</a>, <a href="https://arxiv.org/format/2502.11740">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> Mitigating Visual Knowledge Forgetting in MLLM Instruction-tuning via Modality-decoupled Gradient Descent </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wu%2C+J">Junda Wu</a>, <a href="/search/cs?searchtype=author&query=Xiong%2C+Y">Yuxin Xiong</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xintong Li</a>, <a href="/search/cs?searchtype=author&query=Xia%2C+Y">Yu Xia</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruoyu Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yu Wang</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+T">Tong Yu</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+S">Sungchul Kim</a>, <a href="/search/cs?searchtype=author&query=Rossi%2C+R+A">Ryan A. Rossi</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+L">Lina Yao</a>, <a href="/search/cs?searchtype=author&query=Shang%2C+J">Jingbo Shang</a>, <a href="/search/cs?searchtype=author&query=McAuley%2C+J">Julian McAuley</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="2502.11740v1-abstract-short" style="display: inline;"> Recent MLLMs have shown emerging visual understanding and reasoning abilities after being pre-trained on large-scale multimodal datasets. Unlike pre-training, where MLLMs receive rich visual-text alignment, instruction-tuning is often text-driven with weaker visual supervision, leading to the degradation of pre-trained visual understanding and causing visual forgetting. Existing approaches, such a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11740v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11740v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11740v1-abstract-full" style="display: none;"> Recent MLLMs have shown emerging visual understanding and reasoning abilities after being pre-trained on large-scale multimodal datasets. Unlike pre-training, where MLLMs receive rich visual-text alignment, instruction-tuning is often text-driven with weaker visual supervision, leading to the degradation of pre-trained visual understanding and causing visual forgetting. Existing approaches, such as direct fine-tuning and continual learning methods, fail to explicitly address this issue, often compressing visual representations and prioritizing task alignment over visual retention, which further worsens visual forgetting. To overcome this limitation, we introduce a novel perspective leveraging effective rank to quantify the degradation of visual representation richness, interpreting this degradation through the information bottleneck principle as excessive compression that leads to the degradation of crucial pre-trained visual knowledge. Building on this view, we propose a modality-decoupled gradient descent (MDGD) method that regulates gradient updates to maintain the effective rank of visual representations while mitigating the over-compression effects described by the information bottleneck. By explicitly disentangling the optimization of visual understanding from task-specific alignment, MDGD preserves pre-trained visual knowledge while enabling efficient task adaptation. To enable lightweight instruction-tuning, we further develop a memory-efficient fine-tuning approach using gradient masking, which selectively updates a subset of model parameters to enable parameter-efficient fine-tuning (PEFT), reducing computational overhead while preserving rich visual representations. Extensive experiments across various downstream tasks and backbone MLLMs demonstrate that MDGD effectively mitigates visual forgetting from pre-trained tasks while enabling strong adaptation to new tasks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11740v1-abstract-full').style.display = 'none'; document.getElementById('2502.11740v1-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">9 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/2502.11616">arXiv:2502.11616</a> <span> [<a href="https://arxiv.org/pdf/2502.11616">pdf</a>, <a href="https://arxiv.org/format/2502.11616">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</span> </div> </div> <p class="title is-5 mathjax"> User-Centric Data Management in Decentralized Internet of Behaviors System </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+S">Shiqi Zhang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+D">Dapeng Wu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+H">Honggang Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruyan 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="2502.11616v1-abstract-short" style="display: inline;"> The Internet of Behaviors (IoB) is an emerging concept that utilizes devices to collect human behavior and provide intelligent services. Although some research has focused on human behavior analysis and data collection within IoB, the associated security and privacy challenges remain insufficiently explored. This paper analyzes the security and privacy risks at different stages of behavioral data… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11616v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11616v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11616v1-abstract-full" style="display: none;"> The Internet of Behaviors (IoB) is an emerging concept that utilizes devices to collect human behavior and provide intelligent services. Although some research has focused on human behavior analysis and data collection within IoB, the associated security and privacy challenges remain insufficiently explored. This paper analyzes the security and privacy risks at different stages of behavioral data generating, uploading, and using, while also considering the dynamic characteristics of user activity areas. Then, we propose a blockchain-based distributed IoB data storage and sharing framework, which is categorized into sensing, processing, and management layers based on these stages. To accommodate both identity authentication and behavioral privacy, zero-knowledge proofs are used in the sensing layer to separate the correlation between behavior and identity, which is further extended to a distributed architecture for cross-domain authentication. In the processing layer, an improved consensus protocol is proposed to enhance the decision-making efficiency of distributed IoB by analyzing the geographical and computational capability of the servers. In the management layer, user permission differences and the privacy of access targets are considered. Different types of behavior are modeled as corresponding relationships between keys, and fine-grained secure access is achieved through function secret sharing. Simulation results demonstrate the effectiveness of the proposed framework in multi-scenario IoB, with average consensus and authentication times reduced by 74% and 56%, respectively. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11616v1-abstract-full').style.display = 'none'; document.getElementById('2502.11616v1-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/2502.11563">arXiv:2502.11563</a> <span> [<a href="https://arxiv.org/pdf/2502.11563">pdf</a>, <a href="https://arxiv.org/format/2502.11563">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> </div> </div> <p class="title is-5 mathjax"> Leader and Follower: Interactive Motion Generation under Trajectory Constraints </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runqi Wang</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+C">Caoyuan Ma</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+J">Jian Zhao</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+H">Hanrui Xu</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+D">Dongfang Sun</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+H">Haoyang Chen</a>, <a href="/search/cs?searchtype=author&query=Xiong%2C+L">Lin Xiong</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zheng Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xuelong 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="2502.11563v1-abstract-short" style="display: inline;"> With the rapid advancement of game and film production, generating interactive motion from texts has garnered significant attention due to its potential to revolutionize content creation processes. In many practical applications, there is a need to impose strict constraints on the motion range or trajectory of virtual characters. However, existing methods that rely solely on textual input face sub… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11563v1-abstract-full').style.display = 'inline'; document.getElementById('2502.11563v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.11563v1-abstract-full" style="display: none;"> With the rapid advancement of game and film production, generating interactive motion from texts has garnered significant attention due to its potential to revolutionize content creation processes. In many practical applications, there is a need to impose strict constraints on the motion range or trajectory of virtual characters. However, existing methods that rely solely on textual input face substantial challenges in accurately capturing the user's intent, particularly in specifying the desired trajectory. As a result, the generated motions often lack plausibility and accuracy. Moreover, existing trajectory - based methods for customized motion generation rely on retraining for single - actor scenarios, which limits flexibility and adaptability to different datasets, as well as interactivity in two-actor motions. To generate interactive motion following specified trajectories, this paper decouples complex motion into a Leader - Follower dynamic, inspired by role allocation in partner dancing. Based on this framework, this paper explores the motion range refinement process in interactive motion generation and proposes a training-free approach, integrating a Pace Controller and a Kinematic Synchronization Adapter. The framework enhances the ability of existing models to generate motion that adheres to trajectory by controlling the leader's movement and correcting the follower's motion to align with the leader. Experimental results show that the proposed approach, by better leveraging trajectory information, outperforms existing methods in both realism and accuracy. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.11563v1-abstract-full').style.display = 'none'; document.getElementById('2502.11563v1-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/2502.10841">arXiv:2502.10841</a> <span> [<a href="https://arxiv.org/pdf/2502.10841">pdf</a>, <a href="https://arxiv.org/format/2502.10841">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"> SkyReels-A1: Expressive Portrait Animation in Video Diffusion Transformers </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Qiu%2C+D">Di Qiu</a>, <a href="/search/cs?searchtype=author&query=Fei%2C+Z">Zhengcong Fei</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui Wang</a>, <a href="/search/cs?searchtype=author&query=Bai%2C+J">Jialin Bai</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+C">Changqian Yu</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+M">Mingyuan Fan</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+G">Guibin Chen</a>, <a href="/search/cs?searchtype=author&query=Wen%2C+X">Xiang Wen</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="2502.10841v1-abstract-short" style="display: inline;"> We present SkyReels-A1, a simple yet effective framework built upon video diffusion Transformer to facilitate portrait image animation. Existing methodologies still encounter issues, including identity distortion, background instability, and unrealistic facial dynamics, particularly in head-only animation scenarios. Besides, extending to accommodate diverse body proportions usually leads to visual… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10841v1-abstract-full').style.display = 'inline'; document.getElementById('2502.10841v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.10841v1-abstract-full" style="display: none;"> We present SkyReels-A1, a simple yet effective framework built upon video diffusion Transformer to facilitate portrait image animation. Existing methodologies still encounter issues, including identity distortion, background instability, and unrealistic facial dynamics, particularly in head-only animation scenarios. Besides, extending to accommodate diverse body proportions usually leads to visual inconsistencies or unnatural articulations. To address these challenges, SkyReels-A1 capitalizes on the strong generative capabilities of video DiT, enhancing facial motion transfer precision, identity retention, and temporal coherence. The system incorporates an expression-aware conditioning module that enables seamless video synthesis driven by expression-guided landmark inputs. Integrating the facial image-text alignment module strengthens the fusion of facial attributes with motion trajectories, reinforcing identity preservation. Additionally, SkyReels-A1 incorporates a multi-stage training paradigm to incrementally refine the correlation between expressions and motion while ensuring stable identity reproduction. Extensive empirical evaluations highlight the model's ability to produce visually coherent and compositionally diverse results, making it highly applicable to domains such as virtual avatars, remote communication, and digital media generation. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.10841v1-abstract-full').style.display = 'none'; document.getElementById('2502.10841v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 15 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </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/2502.09797">arXiv:2502.09797</a> <span> [<a href="https://arxiv.org/pdf/2502.09797">pdf</a>, <a href="https://arxiv.org/format/2502.09797">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Information Retrieval">cs.IR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> A Survey on LLM-based News Recommender Systems </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rongyao Wang</a>, <a href="/search/cs?searchtype=author&query=Liesaputra%2C+V">Veronica Liesaputra</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Z">Zhiyi Huang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.09797v2-abstract-short" style="display: inline;"> News recommender systems play a critical role in mitigating the information overload problem. In recent years, due to the successful applications of large language model technologies, researchers have utilized Discriminative Large Language Models (DLLMs) or Generative Large Language Models (GLLMs) to improve the performance of news recommender systems. Although several recent surveys review signif… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09797v2-abstract-full').style.display = 'inline'; document.getElementById('2502.09797v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.09797v2-abstract-full" style="display: none;"> News recommender systems play a critical role in mitigating the information overload problem. In recent years, due to the successful applications of large language model technologies, researchers have utilized Discriminative Large Language Models (DLLMs) or Generative Large Language Models (GLLMs) to improve the performance of news recommender systems. Although several recent surveys review significant challenges for deep learning-based news recommender systems, such as fairness, privacy-preserving, and responsibility, there is a lack of a systematic survey on Large Language Model (LLM)-based news recommender systems. In order to review different core methodologies and explore potential issues systematically, we categorize DLLM-based and GLLM-based news recommender systems under the umbrella of LLM-based news recommender systems. In this survey, we first overview the development of deep learning-based news recommender systems. Then, we review LLM-based news recommender systems based on three aspects: news-oriented modeling, user-oriented modeling, and prediction-oriented modeling. Next, we examine the challenges from various perspectives, including datasets, benchmarking tools, and methodologies. Furthermore, we conduct extensive experiments to analyze how large language model technologies affect the performance of different news recommender systems. Finally, we comprehensively explore the future directions for LLM-based news recommendations in the era of LLMs. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09797v2-abstract-full').style.display = 'none'; document.getElementById('2502.09797v2-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 13 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">20 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/2502.09649">arXiv:2502.09649</a> <span> [<a href="https://arxiv.org/pdf/2502.09649">pdf</a>, <a href="https://arxiv.org/format/2502.09649">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="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> </div> </div> <p class="title is-5 mathjax"> Imit Diff: Semantics Guided Diffusion Transformer with Dual Resolution Fusion for Imitation Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Dong%2C+Y">Yuhang Dong</a>, <a href="/search/cs?searchtype=author&query=Ge%2C+H">Haizhou Ge</a>, <a href="/search/cs?searchtype=author&query=Zeng%2C+Y">Yupei Zeng</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jiangning Zhang</a>, <a href="/search/cs?searchtype=author&query=Tian%2C+B">Beiwen Tian</a>, <a href="/search/cs?searchtype=author&query=Tian%2C+G">Guanzhong Tian</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+H">Hongrui Zhu</a>, <a href="/search/cs?searchtype=author&query=Jia%2C+Y">Yufei Jia</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruixiang Wang</a>, <a href="/search/cs?searchtype=author&query=Yi%2C+R">Ran Yi</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+G">Guyue Zhou</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+L">Longhua Ma</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="2502.09649v1-abstract-short" style="display: inline;"> Visuomotor imitation learning enables embodied agents to effectively acquire manipulation skills from video demonstrations and robot proprioception. However, as scene complexity and visual distractions increase, existing methods that perform well in simple scenes tend to degrade in performance. To address this challenge, we introduce Imit Diff, a semanstic guided diffusion transformer with dual re… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09649v1-abstract-full').style.display = 'inline'; document.getElementById('2502.09649v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.09649v1-abstract-full" style="display: none;"> Visuomotor imitation learning enables embodied agents to effectively acquire manipulation skills from video demonstrations and robot proprioception. However, as scene complexity and visual distractions increase, existing methods that perform well in simple scenes tend to degrade in performance. To address this challenge, we introduce Imit Diff, a semanstic guided diffusion transformer with dual resolution fusion for imitation learning. Our approach leverages prior knowledge from vision language foundation models to translate high-level semantic instruction into pixel-level visual localization. This information is explicitly integrated into a multi-scale visual enhancement framework, constructed with a dual resolution encoder. Additionally, we introduce an implementation of Consistency Policy within the diffusion transformer architecture to improve both real-time performance and motion smoothness in embodied agent control.We evaluate Imit Diff on several challenging real-world tasks. Due to its task-oriented visual localization and fine-grained scene perception, it significantly outperforms state-of-the-art methods, especially in complex scenes with visual distractions, including zero-shot experiments focused on visual distraction and category generalization. The code will be made publicly available. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09649v1-abstract-full').style.display = 'none'; document.getElementById('2502.09649v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 11 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.09571">arXiv:2502.09571</a> <span> [<a href="https://arxiv.org/pdf/2502.09571">pdf</a>, <a href="https://arxiv.org/format/2502.09571">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="Quantitative Methods">q-bio.QM</span> </div> </div> <p class="title is-5 mathjax"> DiffMS: Diffusion Generation of Molecules Conditioned on Mass Spectra </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Bohde%2C+M">Montgomery Bohde</a>, <a href="/search/cs?searchtype=author&query=Manjrekar%2C+M">Mrunali Manjrekar</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzhong Wang</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+S">Shuiwang Ji</a>, <a href="/search/cs?searchtype=author&query=Coley%2C+C+W">Connor W. Coley</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="2502.09571v1-abstract-short" style="display: inline;"> Mass spectrometry plays a fundamental role in elucidating the structures of unknown molecules and subsequent scientific discoveries. One formulation of the structure elucidation task is the conditional $\textit{de novo}$ generation of molecular structure given a mass spectrum. Toward a more accurate and efficient scientific discovery pipeline for small molecules, we present DiffMS, a formula-restr… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09571v1-abstract-full').style.display = 'inline'; document.getElementById('2502.09571v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.09571v1-abstract-full" style="display: none;"> Mass spectrometry plays a fundamental role in elucidating the structures of unknown molecules and subsequent scientific discoveries. One formulation of the structure elucidation task is the conditional $\textit{de novo}$ generation of molecular structure given a mass spectrum. Toward a more accurate and efficient scientific discovery pipeline for small molecules, we present DiffMS, a formula-restricted encoder-decoder generative network that achieves state-of-the-art performance on this task. The encoder utilizes a transformer architecture and models mass spectra domain knowledge such as peak formulae and neutral losses, and the decoder is a discrete graph diffusion model restricted by the heavy-atom composition of a known chemical formula. To develop a robust decoder that bridges latent embeddings and molecular structures, we pretrain the diffusion decoder with fingerprint-structure pairs, which are available in virtually infinite quantities, compared to structure-spectrum pairs that number in the tens of thousands. Extensive experiments on established benchmarks show that DiffMS outperforms existing models on $\textit{de novo}$ molecule generation. We provide several ablations to demonstrate the effectiveness of our diffusion and pretraining approaches and show consistent performance scaling with increasing pretraining dataset size. DiffMS code is publicly available at https://github.com/coleygroup/DiffMS. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.09571v1-abstract-full').style.display = 'none'; document.getElementById('2502.09571v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 13 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Preprint</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.08966">arXiv:2502.08966</a> <span> [<a href="https://arxiv.org/pdf/2502.08966">pdf</a>, <a href="https://arxiv.org/format/2502.08966">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> RTBAS: Defending LLM Agents Against Prompt Injection and Privacy Leakage </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhong%2C+P+Y">Peter Yong Zhong</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+S">Siyuan Chen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruiqi Wang</a>, <a href="/search/cs?searchtype=author&query=McCall%2C+M">McKenna McCall</a>, <a href="/search/cs?searchtype=author&query=Titzer%2C+B+L">Ben L. Titzer</a>, <a href="/search/cs?searchtype=author&query=Miller%2C+H">Heather Miller</a>, <a href="/search/cs?searchtype=author&query=Gibbons%2C+P+B">Phillip B. Gibbons</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="2502.08966v2-abstract-short" style="display: inline;"> Tool-Based Agent Systems (TBAS) allow Language Models (LMs) to use external tools for tasks beyond their standalone capabilities, such as searching websites, booking flights, or making financial transactions. However, these tools greatly increase the risks of prompt injection attacks, where malicious content hijacks the LM agent to leak confidential data or trigger harmful actions. Existing defens… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08966v2-abstract-full').style.display = 'inline'; document.getElementById('2502.08966v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08966v2-abstract-full" style="display: none;"> Tool-Based Agent Systems (TBAS) allow Language Models (LMs) to use external tools for tasks beyond their standalone capabilities, such as searching websites, booking flights, or making financial transactions. However, these tools greatly increase the risks of prompt injection attacks, where malicious content hijacks the LM agent to leak confidential data or trigger harmful actions. Existing defenses (OpenAI GPTs) require user confirmation before every tool call, placing onerous burdens on users. We introduce Robust TBAS (RTBAS), which automatically detects and executes tool calls that preserve integrity and confidentiality, requiring user confirmation only when these safeguards cannot be ensured. RTBAS adapts Information Flow Control to the unique challenges presented by TBAS. We present two novel dependency screeners, using LM-as-a-judge and attention-based saliency, to overcome these challenges. Experimental results on the AgentDojo Prompt Injection benchmark show RTBAS prevents all targeted attacks with only a 2% loss of task utility when under attack, and further tests confirm its ability to obtain near-oracle performance on detecting both subtle and direct privacy leaks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08966v2-abstract-full').style.display = 'none'; document.getElementById('2502.08966v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 13 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 13 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/2502.08504">arXiv:2502.08504</a> <span> [<a href="https://arxiv.org/pdf/2502.08504">pdf</a>, <a href="https://arxiv.org/format/2502.08504">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"> MoDitector: Module-Directed Testing for Autonomous Driving Systems </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Renzhi Wang</a>, <a href="/search/cs?searchtype=author&query=Cheng%2C+M">Mingfei Cheng</a>, <a href="/search/cs?searchtype=author&query=Xie%2C+X">Xiaofei Xie</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Y">Yuan Zhou</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+L">Lei Ma</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="2502.08504v1-abstract-short" style="display: inline;"> Testing Autonomous Driving Systems (ADS) is crucial for ensuring their safety, reliability, and performance. Despite numerous testing methods available that can generate diverse and challenging scenarios to uncover potential vulnerabilities, these methods often treat ADS as a black-box, primarily focusing on identifying system failures like collisions or near-misses without pinpointing the specifi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08504v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08504v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08504v1-abstract-full" style="display: none;"> Testing Autonomous Driving Systems (ADS) is crucial for ensuring their safety, reliability, and performance. Despite numerous testing methods available that can generate diverse and challenging scenarios to uncover potential vulnerabilities, these methods often treat ADS as a black-box, primarily focusing on identifying system failures like collisions or near-misses without pinpointing the specific modules responsible for these failures. Understanding the root causes of failures is essential for effective debugging and subsequent system repair. We observed that existing methods also fall short in generating diverse failures that adequately test the distinct modules of an ADS, such as perception, prediction, planning and control. To bridge this gap, we introduce MoDitector, the first root-cause-aware testing method for ADS. Unlike previous approaches, MoDitector not only generates scenarios leading to collisions but also showing which specific module triggered the failure. This method targets specific modules, creating test scenarios that highlight the weaknesses of these given modules. Specifically, our approach involves designing module-specific oracles to ascertain module failures and employs a module-directed testing strategy that includes module-specific feedback, adaptive seed selection, and mutation. This strategy guides the generation of tests that effectively provoke module-specific failures. We evaluated MoDitector across four critical ADS modules and four testing scenarios. Our approach represents a significant innovation in ADS testing by focusing on identifying and rectifying module-specific errors within the system, moving beyond conventional black-box failure detection. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08504v1-abstract-full').style.display = 'none'; document.getElementById('2502.08504v1-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/2502.08161">arXiv:2502.08161</a> <span> [<a href="https://arxiv.org/pdf/2502.08161">pdf</a>, <a href="https://arxiv.org/format/2502.08161">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Information Retrieval">cs.IR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> <div class="is-inline-block" style="margin-left: 0.5rem"> <div class="tags has-addons"> <span class="tag is-dark is-size-7">doi</span> <span class="tag is-light is-size-7"><a class="" href="https://doi.org/10.1109/ICDM54844.2022.00070">10.1109/ICDM54844.2022.00070 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> MixDec Sampling: A Soft Link-based Sampling Method of Graph Neural Network for Recommendation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Xie%2C+X">Xiangjin Xie</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yuxin Chen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruipeng Wang</a>, <a href="/search/cs?searchtype=author&query=Ouyang%2C+K">Kai Ouyang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zihan Zhang</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+H">Hai-Tao Zheng</a>, <a href="/search/cs?searchtype=author&query=Qian%2C+B">Buyue Qian</a>, <a href="/search/cs?searchtype=author&query=Zheng%2C+H">Hansen Zheng</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+B">Bo Hu</a>, <a href="/search/cs?searchtype=author&query=Zhuo%2C+C">Chengxiang Zhuo</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Z">Zang 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="2502.08161v1-abstract-short" style="display: inline;"> Graph neural networks have been widely used in recent recommender systems, where negative sampling plays an important role. Existing negative sampling methods restrict the relationship between nodes as either hard positive pairs or hard negative pairs. This leads to the loss of structural information, and lacks the mechanism to generate positive pairs for nodes with few neighbors. To overcome limi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08161v1-abstract-full').style.display = 'inline'; document.getElementById('2502.08161v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.08161v1-abstract-full" style="display: none;"> Graph neural networks have been widely used in recent recommender systems, where negative sampling plays an important role. Existing negative sampling methods restrict the relationship between nodes as either hard positive pairs or hard negative pairs. This leads to the loss of structural information, and lacks the mechanism to generate positive pairs for nodes with few neighbors. To overcome limitations, we propose a novel soft link-based sampling method, namely MixDec Sampling, which consists of Mixup Sampling module and Decay Sampling module. The Mixup Sampling augments node features by synthesizing new nodes and soft links, which provides sufficient number of samples for nodes with few neighbors. The Decay Sampling strengthens the digestion of graph structure information by generating soft links for node embedding learning. To the best of our knowledge, we are the first to model sampling relationships between nodes by soft links in GNN-based recommender systems. Extensive experiments demonstrate that the proposed MixDec Sampling can significantly and consistently improve the recommendation performance of several representative GNN-based models on various recommendation benchmarks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.08161v1-abstract-full').style.display = 'none'; document.getElementById('2502.08161v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 12 February, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> February 2025. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">10 pages, 6 figures</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06832">arXiv:2502.06832</a> <span> [<a href="https://arxiv.org/pdf/2502.06832">pdf</a>, <a href="https://arxiv.org/format/2502.06832">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"> Optimizing Robustness and Accuracy in Mixture of Experts: A Dual-Model Approach </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+X">Xu Zhang</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+K">Kaidi Xu</a>, <a href="/search/cs?searchtype=author&query=Hu%2C+Z">Ziqing Hu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ren 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="2502.06832v2-abstract-short" style="display: inline;"> Mixture of Experts (MoE) have shown remarkable success in leveraging specialized expert networks for complex machine learning tasks. However, their susceptibility to adversarial attacks presents a critical challenge for deployment in robust applications. This paper addresses the critical question of how to incorporate robustness into MoEs while maintaining high natural accuracy. We begin by analyz… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06832v2-abstract-full').style.display = 'inline'; document.getElementById('2502.06832v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06832v2-abstract-full" style="display: none;"> Mixture of Experts (MoE) have shown remarkable success in leveraging specialized expert networks for complex machine learning tasks. However, their susceptibility to adversarial attacks presents a critical challenge for deployment in robust applications. This paper addresses the critical question of how to incorporate robustness into MoEs while maintaining high natural accuracy. We begin by analyzing the vulnerability of MoE components, finding that expert networks are notably more susceptible to adversarial attacks than the router. Based on this insight, we propose a targeted robust training technique that integrates a novel loss function to enhance the adversarial robustness of MoE, requiring only the robustification of one additional expert without compromising training or inference efficiency. Building on this, we introduce a dual-model strategy that linearly combines a standard MoE model with our robustified MoE model using a smoothing parameter. This approach allows for flexible control over the robustness-accuracy trade-off. We further provide theoretical foundations by deriving certified robustness bounds for both the single MoE and the dual-model. To push the boundaries of robustness and accuracy, we propose a novel joint training strategy JTDMoE for the dual-model. This joint training enhances both robustness and accuracy beyond what is achievable with separate models. Experimental results on CIFAR-10 and TinyImageNet datasets using ResNet18 and Vision Transformer (ViT) architectures demonstrate the effectiveness of our proposed methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06832v2-abstract-full').style.display = 'none'; document.getElementById('2502.06832v2-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 5 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">10 pages, 3 figures, submitted to ICML 2025 (under review)</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.2.6; I.5.1 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06650">arXiv:2502.06650</a> <span> [<a href="https://arxiv.org/pdf/2502.06650">pdf</a>, <a href="https://arxiv.org/format/2502.06650">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"> Prototype Contrastive Consistency Learning for Semi-Supervised Medical Image Segmentation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=He%2C+S">Shihuan He</a>, <a href="/search/cs?searchtype=author&query=Lai%2C+Z">Zhihui Lai</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruxin Wang</a>, <a href="/search/cs?searchtype=author&query=Kong%2C+H">Heng Kong</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="2502.06650v1-abstract-short" style="display: inline;"> Medical image segmentation is a crucial task in medical image analysis, but it can be very challenging especially when there are less labeled data but with large unlabeled data. Contrastive learning has proven to be effective for medical image segmentation in semi-supervised learning by constructing contrastive samples from partial pixels. However, although previous contrastive learning methods ca… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06650v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06650v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06650v1-abstract-full" style="display: none;"> Medical image segmentation is a crucial task in medical image analysis, but it can be very challenging especially when there are less labeled data but with large unlabeled data. Contrastive learning has proven to be effective for medical image segmentation in semi-supervised learning by constructing contrastive samples from partial pixels. However, although previous contrastive learning methods can mine semantic information from partial pixels within images, they ignore the whole context information of unlabeled images, which is very important to precise segmentation. In order to solve this problem, we propose a novel prototype contrastive learning method called Prototype Contrastive Consistency Segmentation (PCCS) for semi-supervised medical image segmentation. The core idea is to enforce the prototypes of the same semantic class to be closer and push the prototypes in different semantic classes far away from each other. Specifically, we construct a signed distance map and an uncertainty map from unlabeled images. The signed distance map is used to construct prototypes for contrastive learning, and then we estimate the prototype uncertainty from the uncertainty map as trade-off among prototypes. In order to obtain better prototypes, based on the student-teacher architecture, a new mechanism named prototype updating prototype is designed to assist in updating the prototypes for contrastive learning. In addition, we propose an uncertainty-consistency loss to mine more reliable information from unlabeled data. Extensive experiments on medical image segmentation demonstrate that PCCS achieves better segmentation performance than the state-of-the-art methods. The code is available at https://github.com/comphsh/PCCS. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06650v1-abstract-full').style.display = 'none'; document.getElementById('2502.06650v1-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">17 pages, 10 figures, 7 tables</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.4.6; I.5.4 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06589">arXiv:2502.06589</a> <span> [<a href="https://arxiv.org/pdf/2502.06589">pdf</a>, <a href="https://arxiv.org/format/2502.06589">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Hephaestus: Improving Fundamental Agent Capabilities of Large Language Models through Continual Pre-Training </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhuang%2C+Y">Yuchen Zhuang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+J">Jingfeng Yang</a>, <a href="/search/cs?searchtype=author&query=Jiang%2C+H">Haoming Jiang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xin Liu</a>, <a href="/search/cs?searchtype=author&query=Cheng%2C+K">Kewei Cheng</a>, <a href="/search/cs?searchtype=author&query=Lokegaonkar%2C+S">Sanket Lokegaonkar</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+Y">Yifan Gao</a>, <a href="/search/cs?searchtype=author&query=Ping%2C+Q">Qing Ping</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+T">Tianyi Liu</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+B">Binxuan Huang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Z">Zheng Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhengyang Wang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+P">Pei Chen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruijie Wang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+R">Rongzhi Zhang</a>, <a href="/search/cs?searchtype=author&query=Zalmout%2C+N">Nasser Zalmout</a>, <a href="/search/cs?searchtype=author&query=Nigam%2C+P">Priyanka Nigam</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+B">Bing Yin</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chao 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="2502.06589v1-abstract-short" style="display: inline;"> Due to the scarcity of agent-oriented pre-training data, LLM-based autonomous agents typically rely on complex prompting or extensive fine-tuning, which often fails to introduce new capabilities while preserving strong generalizability. We introduce Hephaestus-Forge, the first large-scale pre-training corpus designed to enhance the fundamental capabilities of LLM agents in API function calling, in… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06589v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06589v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06589v1-abstract-full" style="display: none;"> Due to the scarcity of agent-oriented pre-training data, LLM-based autonomous agents typically rely on complex prompting or extensive fine-tuning, which often fails to introduce new capabilities while preserving strong generalizability. We introduce Hephaestus-Forge, the first large-scale pre-training corpus designed to enhance the fundamental capabilities of LLM agents in API function calling, intrinsic reasoning and planning, and adapting to environmental feedback. Hephaestus-Forge comprises 103B agent-specific data encompassing 76,537 APIs, including both tool documentation to introduce knowledge of API functions and function calling trajectories to strengthen intrinsic reasoning. To explore effective training protocols, we investigate scaling laws to identify the optimal recipe in data mixing ratios. By continual pre-training on Hephaestus-Forge, Hephaestus outperforms small- to medium-scale open-source LLMs and rivals commercial LLMs on three agent benchmarks, demonstrating the effectiveness of our pre-training corpus in enhancing fundamental agentic capabilities and generalization of LLMs to new tasks or environments. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06589v1-abstract-full').style.display = 'none'; document.getElementById('2502.06589v1-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 to NAACL 2025 main conference</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06453">arXiv:2502.06453</a> <span> [<a href="https://arxiv.org/pdf/2502.06453">pdf</a>, <a href="https://arxiv.org/format/2502.06453">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"> MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Huang%2C+K">Kaixuan Huang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+J">Jiacheng Guo</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Z">Zihao Li</a>, <a href="/search/cs?searchtype=author&query=Ji%2C+X">Xiang Ji</a>, <a href="/search/cs?searchtype=author&query=Ge%2C+J">Jiawei Ge</a>, <a href="/search/cs?searchtype=author&query=Li%2C+W">Wenzhe Li</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+Y">Yingqing Guo</a>, <a href="/search/cs?searchtype=author&query=Cai%2C+T">Tianle Cai</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+H">Hui Yuan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzhe Wang</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+Y">Yue Wu</a>, <a href="/search/cs?searchtype=author&query=Yin%2C+M">Ming Yin</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+S">Shange Tang</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yangsibo Huang</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+C">Chi Jin</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xinyun Chen</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+M">Mengdi 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="2502.06453v2-abstract-short" style="display: inline;"> Large language models have demonstrated impressive performance on challenging mathematical reasoning tasks, which has triggered the discussion of whether the performance is achieved by true reasoning capability or memorization. To investigate this question, prior work has constructed mathematical benchmarks when questions undergo simple perturbations -- modifications that still preserve the underl… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06453v2-abstract-full').style.display = 'inline'; document.getElementById('2502.06453v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06453v2-abstract-full" style="display: none;"> Large language models have demonstrated impressive performance on challenging mathematical reasoning tasks, which has triggered the discussion of whether the performance is achieved by true reasoning capability or memorization. To investigate this question, prior work has constructed mathematical benchmarks when questions undergo simple perturbations -- modifications that still preserve the underlying reasoning patterns of the solutions. However, no work has explored hard perturbations, which fundamentally change the nature of the problem so that the original solution steps do not apply. To bridge the gap, we construct MATH-P-Simple and MATH-P-Hard via simple perturbation and hard perturbation, respectively. Each consists of 279 perturbed math problems derived from level-5 (hardest) problems in the MATH dataset (Hendrycksmath et. al., 2021). We observe significant performance drops on MATH-P-Hard across various models, including o1-mini (-16.49%) and gemini-2.0-flash-thinking (-12.9%). We also raise concerns about a novel form of memorization where models blindly apply learned problem-solving skills without assessing their applicability to modified contexts. This issue is amplified when using original problems for in-context learning. We call for research efforts to address this challenge, which is critical for developing more robust and reliable reasoning models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06453v2-abstract-full').style.display = 'none'; document.getElementById('2502.06453v2-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 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">v2: fix bugs in Fig. 1</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06274">arXiv:2502.06274</a> <span> [<a href="https://arxiv.org/pdf/2502.06274">pdf</a>, <a href="https://arxiv.org/format/2502.06274">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="Molecular Networks">q-bio.MN</span> </div> </div> <p class="title is-5 mathjax"> HODDI: A Dataset of High-Order Drug-Drug Interactions for Computational Pharmacovigilance </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Z">Zhaoying Wang</a>, <a href="/search/cs?searchtype=author&query=Shi%2C+Y">Yingdan Shi</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xiang Liu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+C">Can Chen</a>, <a href="/search/cs?searchtype=author&query=Wen%2C+J">Jun Wen</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ren 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="2502.06274v1-abstract-short" style="display: inline;"> Drug-side effect research is vital for understanding adverse reactions arising in complex multi-drug therapies. However, the scarcity of higher-order datasets that capture the combinatorial effects of multiple drugs severely limits progress in this field. Existing resources such as TWOSIDES primarily focus on pairwise interactions. To fill this critical gap, we introduce HODDI, the first Higher-Or… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06274v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06274v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06274v1-abstract-full" style="display: none;"> Drug-side effect research is vital for understanding adverse reactions arising in complex multi-drug therapies. However, the scarcity of higher-order datasets that capture the combinatorial effects of multiple drugs severely limits progress in this field. Existing resources such as TWOSIDES primarily focus on pairwise interactions. To fill this critical gap, we introduce HODDI, the first Higher-Order Drug-Drug Interaction Dataset, constructed from U.S. Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) records spanning the past decade, to advance computational pharmacovigilance. HODDI contains 109,744 records involving 2,506 unique drugs and 4,569 unique side effects, specifically curated to capture multi-drug interactions and their collective impact on adverse effects. Comprehensive statistical analyses demonstrate HODDI's extensive coverage and robust analytical metrics, making it a valuable resource for studying higher-order drug relationships. Evaluating HODDI with multiple models, we found that simple Multi-Layer Perceptron (MLP) can outperform graph models, while hypergraph models demonstrate superior performance in capturing complex multi-drug interactions, further validating HODDI's effectiveness. Our findings highlight the inherent value of higher-order information in drug-side effect prediction and position HODDI as a benchmark dataset for advancing research in pharmacovigilance, drug safety, and personalized medicine. The dataset and codes are available at https://github.com/TIML-Group/HODDI. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06274v1-abstract-full').style.display = 'none'; document.getElementById('2502.06274v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.06193">arXiv:2502.06193</a> <span> [<a href="https://arxiv.org/pdf/2502.06193">pdf</a>, <a href="https://arxiv.org/format/2502.06193">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> <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"> Can LLMs Replace Human Evaluators? An Empirical Study of LLM-as-a-Judge in Software Engineering </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruiqi Wang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+J">Jiyu Guo</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+C">Cuiyun Gao</a>, <a href="/search/cs?searchtype=author&query=Fan%2C+G">Guodong Fan</a>, <a href="/search/cs?searchtype=author&query=Chong%2C+C+Y">Chun Yong Chong</a>, <a href="/search/cs?searchtype=author&query=Xia%2C+X">Xin Xia</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="2502.06193v1-abstract-short" style="display: inline;"> Recently, large language models (LLMs) have been deployed to tackle various software engineering (SE) tasks like code generation, significantly advancing the automation of SE tasks. However, assessing the quality of these LLM-generated code and text remains challenging. The commonly used Pass@k metric necessitates extensive unit tests and configured environments, demands a high labor cost, and is… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06193v1-abstract-full').style.display = 'inline'; document.getElementById('2502.06193v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06193v1-abstract-full" style="display: none;"> Recently, large language models (LLMs) have been deployed to tackle various software engineering (SE) tasks like code generation, significantly advancing the automation of SE tasks. However, assessing the quality of these LLM-generated code and text remains challenging. The commonly used Pass@k metric necessitates extensive unit tests and configured environments, demands a high labor cost, and is not suitable for evaluating LLM-generated text. Conventional metrics like BLEU, which measure only lexical rather than semantic similarity, have also come under scrutiny. In response, a new trend has emerged to employ LLMs for automated evaluation, known as LLM-as-a-judge. These LLM-as-a-judge methods are claimed to better mimic human assessment than conventional metrics without relying on high-quality reference answers. Nevertheless, their exact human alignment in SE tasks remains unexplored. In this paper, we empirically explore LLM-as-a-judge methods for evaluating SE tasks, focusing on their alignment with human judgments. We select seven LLM-as-a-judge methods that utilize general-purpose LLMs, alongside two LLMs specifically fine-tuned for evaluation. After generating and manually scoring LLM responses on three recent SE datasets of code translation, code generation, and code summarization, we then prompt these methods to evaluate each response. Finally, we compare the scores generated by these methods with human evaluation. The results indicate that output-based methods reach the highest Pearson correlation of 81.32 and 68.51 with human scores in code translation and generation, achieving near-human evaluation, noticeably outperforming ChrF++, one of the best conventional metrics, at 34.23 and 64.92. Such output-based methods prompt LLMs to output judgments directly, and exhibit more balanced score distributions that resemble human score patterns. Finally, we provide... <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06193v1-abstract-full').style.display = 'none'; document.getElementById('2502.06193v1-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 ISSTA 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/2502.06111">arXiv:2502.06111</a> <span> [<a href="https://arxiv.org/pdf/2502.06111">pdf</a>, <a href="https://arxiv.org/format/2502.06111">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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> CSR-Bench: Benchmarking LLM Agents in Deployment of Computer Science Research Repositories </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Xiao%2C+Y">Yijia Xiao</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runhui Wang</a>, <a href="/search/cs?searchtype=author&query=Kong%2C+L">Luyang Kong</a>, <a href="/search/cs?searchtype=author&query=Golac%2C+D">Davor Golac</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+W">Wei 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="2502.06111v2-abstract-short" style="display: inline;"> The increasing complexity of computer science research projects demands more effective tools for deploying code repositories. Large Language Models (LLMs), such as Anthropic Claude and Meta Llama, have demonstrated significant advancements across various fields of computer science research, including the automation of diverse software engineering tasks. To evaluate the effectiveness of LLMs in han… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06111v2-abstract-full').style.display = 'inline'; document.getElementById('2502.06111v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.06111v2-abstract-full" style="display: none;"> The increasing complexity of computer science research projects demands more effective tools for deploying code repositories. Large Language Models (LLMs), such as Anthropic Claude and Meta Llama, have demonstrated significant advancements across various fields of computer science research, including the automation of diverse software engineering tasks. To evaluate the effectiveness of LLMs in handling complex code development tasks of research projects, particularly for NLP/CV/AI/ML/DM topics, we introduce CSR-Bench, a benchmark for Computer Science Research projects. This benchmark assesses LLMs from various aspects including accuracy, efficiency, and deployment script quality, aiming to explore their potential in conducting computer science research autonomously. We also introduce a novel framework, CSR-Agents, that utilizes multiple LLM agents to automate the deployment of GitHub code repositories of computer science research projects. Specifically, by checking instructions from markdown files and interpreting repository structures, the model generates and iteratively improves bash commands that set up the experimental environments and deploy the code to conduct research tasks. Preliminary results from CSR-Bench indicate that LLM agents can significantly enhance the workflow of repository deployment, thereby boosting developer productivity and improving the management of developmental workflows. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.06111v2-abstract-full').style.display = 'none'; document.getElementById('2502.06111v2-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 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/2502.05907">arXiv:2502.05907</a> <span> [<a href="https://arxiv.org/pdf/2502.05907">pdf</a>, <a href="https://arxiv.org/format/2502.05907">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"> EvoAgent: Agent Autonomous Evolution with Continual World Model for Long-Horizon Tasks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Feng%2C+T">Tongtong Feng</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+X">Xin Wang</a>, <a href="/search/cs?searchtype=author&query=Zhou%2C+Z">Zekai Zhou</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ren Wang</a>, <a href="/search/cs?searchtype=author&query=Zhan%2C+Y">Yuwei Zhan</a>, <a href="/search/cs?searchtype=author&query=Li%2C+G">Guangyao Li</a>, <a href="/search/cs?searchtype=author&query=Li%2C+Q">Qing Li</a>, <a href="/search/cs?searchtype=author&query=Zhu%2C+W">Wenwu Zhu</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="2502.05907v1-abstract-short" style="display: inline;"> Completing Long-Horizon (LH) tasks in open-ended worlds is an important yet difficult problem for embodied agents. Existing approaches suffer from two key challenges: (1) they heavily rely on experiences obtained from human-created data or curricula, lacking the ability to continuously update multimodal experiences, and (2) they may encounter catastrophic forgetting issues when faced with new task… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05907v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05907v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05907v1-abstract-full" style="display: none;"> Completing Long-Horizon (LH) tasks in open-ended worlds is an important yet difficult problem for embodied agents. Existing approaches suffer from two key challenges: (1) they heavily rely on experiences obtained from human-created data or curricula, lacking the ability to continuously update multimodal experiences, and (2) they may encounter catastrophic forgetting issues when faced with new tasks, lacking the ability to continuously update world knowledge. To solve these challenges, this paper presents EvoAgent, an autonomous-evolving agent with a continual World Model (WM), which can autonomously complete various LH tasks across environments through self-planning, self-control, and self-reflection, without human intervention. Our proposed EvoAgent contains three modules, i.e., i) the memory-driven planner which uses an LLM along with the WM and interaction memory, to convert LH tasks into executable sub-tasks; ii) the WM-guided action controller which leverages WM to generate low-level actions and incorporates a self-verification mechanism to update multimodal experiences; iii) the experience-inspired reflector which implements a two-stage curriculum learning algorithm to select experiences for task-adaptive WM updates. Moreover, we develop a continual World Model for EvoAgent, which can continuously update the multimodal experience pool and world knowledge through closed-loop dynamics. We conducted extensive experiments on Minecraft, compared with existing methods, EvoAgent can achieve an average success rate improvement of 105% and reduce ineffective actions by more than 6x. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05907v1-abstract-full').style.display = 'none'; document.getElementById('2502.05907v1-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/2502.05206">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/2502.05130">arXiv:2502.05130</a> <span> [<a href="https://arxiv.org/pdf/2502.05130">pdf</a>, <a href="https://arxiv.org/format/2502.05130">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Sound">cs.SD</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Multimedia">cs.MM</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</span> </div> </div> <p class="title is-5 mathjax"> Latent Swap Joint Diffusion for Long-Form Audio Generation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Dai%2C+Y">Yusheng Dai</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+C">Chenxi Wang</a>, <a href="/search/cs?searchtype=author&query=Li%2C+C">Chang Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+C">Chen Wang</a>, <a href="/search/cs?searchtype=author&query=Du%2C+J">Jun Du</a>, <a href="/search/cs?searchtype=author&query=Li%2C+K">Kewei Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruoyu Wang</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+J">Jiefeng Ma</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+L">Lei Sun</a>, <a href="/search/cs?searchtype=author&query=Gao%2C+J">Jianqing Gao</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="2502.05130v1-abstract-short" style="display: inline;"> Previous work on long-form audio generation using global-view diffusion or iterative generation demands significant training or inference costs. While recent advancements in multi-view joint diffusion for panoramic generation provide an efficient option, they struggle with spectrum generation with severe overlap distortions and high cross-view consistency costs. We initially explore this phenomeno… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05130v1-abstract-full').style.display = 'inline'; document.getElementById('2502.05130v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.05130v1-abstract-full" style="display: none;"> Previous work on long-form audio generation using global-view diffusion or iterative generation demands significant training or inference costs. While recent advancements in multi-view joint diffusion for panoramic generation provide an efficient option, they struggle with spectrum generation with severe overlap distortions and high cross-view consistency costs. We initially explore this phenomenon through the connectivity inheritance of latent maps and uncover that averaging operations excessively smooth the high-frequency components of the latent map. To address these issues, we propose Swap Forward (SaFa), a frame-level latent swap framework that synchronizes multiple diffusions to produce a globally coherent long audio with more spectrum details in a forward-only manner. At its core, the bidirectional Self-Loop Latent Swap is applied between adjacent views, leveraging stepwise diffusion trajectory to adaptively enhance high-frequency components without disrupting low-frequency components. Furthermore, to ensure cross-view consistency, the unidirectional Reference-Guided Latent Swap is applied between the reference and the non-overlap regions of each subview during the early stages, providing centralized trajectory guidance. Quantitative and qualitative experiments demonstrate that SaFa significantly outperforms existing joint diffusion methods and even training-based long audio generation models. Moreover, we find that it also adapts well to panoramic generation, achieving comparable state-of-the-art performance with greater efficiency and model generalizability. Project page is available at https://swapforward.github.io/. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.05130v1-abstract-full').style.display = 'none'; document.getElementById('2502.05130v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.04760">arXiv:2502.04760</a> <span> [<a href="https://arxiv.org/pdf/2502.04760">pdf</a>, <a href="https://arxiv.org/format/2502.04760">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"> Graph Federated Learning Based Proactive Content Caching in Edge Computing </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui 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="2502.04760v1-abstract-short" style="display: inline;"> With the rapid growth of mobile data traffic and the increasing prevalence of video streaming, proactive content caching in edge computing has become crucial for reducing latency and alleviating network congestion. However, traditional caching strategies such as FIFO, LRU, and LFU fail to effectively predict future content popularity, while existing proactive caching approaches often require users… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04760v1-abstract-full').style.display = 'inline'; document.getElementById('2502.04760v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.04760v1-abstract-full" style="display: none;"> With the rapid growth of mobile data traffic and the increasing prevalence of video streaming, proactive content caching in edge computing has become crucial for reducing latency and alleviating network congestion. However, traditional caching strategies such as FIFO, LRU, and LFU fail to effectively predict future content popularity, while existing proactive caching approaches often require users to upload data to a central server, raising concerns regarding privacy and scalability. To address these challenges, this paper proposes a Graph Federated Learning-based Proactive Content Caching (GFPCC) scheme that enhances caching efficiency while preserving user privacy. The proposed approach integrates federated learning and graph neural networks, enabling users to locally train Light Graph Convolutional Networks (LightGCN) to capture user-item relationships and predict content popularity. Instead of sharing raw data, only the trained model parameters are transmitted to the central server, where a federated averaging algorithm aggregates updates, refines the global model, and selects the most popular files for proactive caching. Experimental evaluations on real-world datasets, such as MovieLens, demonstrate that GFPCC outperforms baseline caching algorithms by achieving higher cache efficiency through more accurate content popularity predictions. Moreover, the federated learning framework strengthens privacy protection while maintaining efficient model training; however, scalability remains a challenge in large-scale networks with dynamic user preferences. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.04760v1-abstract-full').style.display = 'none'; document.getElementById('2502.04760v1-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> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.03766">arXiv:2502.03766</a> <span> [<a href="https://arxiv.org/pdf/2502.03766">pdf</a>, <a href="https://arxiv.org/ps/2502.03766">ps</a>, <a href="https://arxiv.org/format/2502.03766">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"> Hierarchical Contextual Manifold Alignment for Structuring Latent Representations in Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Dong%2C+M">Meiquan Dong</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+H">Haoran Liu</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yan Huang</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+Z">Zixuan Feng</a>, <a href="/search/cs?searchtype=author&query=Tang%2C+J">Jianhong Tang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruoxi 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="2502.03766v1-abstract-short" style="display: inline;"> The organization of latent token representations plays a crucial role in determining the stability, generalization, and contextual consistency of language models, yet conventional approaches to embedding refinement often rely on parameter modifications that introduce additional computational overhead. A hierarchical alignment method was introduced to restructure token embeddings without altering c… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03766v1-abstract-full').style.display = 'inline'; document.getElementById('2502.03766v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.03766v1-abstract-full" style="display: none;"> The organization of latent token representations plays a crucial role in determining the stability, generalization, and contextual consistency of language models, yet conventional approaches to embedding refinement often rely on parameter modifications that introduce additional computational overhead. A hierarchical alignment method was introduced to restructure token embeddings without altering core model weights, ensuring that representational distributions maintained coherence across different linguistic contexts. Experimental evaluations demonstrated improvements in rare token retrieval, adversarial robustness, and long-range dependency tracking, highlighting the advantages of hierarchical structuring in mitigating inconsistencies in latent space organization. The comparative analysis against conventional fine-tuning and embedding perturbation methods revealed that hierarchical restructuring maintained computational efficiency while achieving measurable gains in representation quality. Structural refinements introduced through the alignment process resulted in improved contextual stability across varied linguistic tasks, reducing inconsistencies in token proximity relationships and enhancing interpretability in language generation. A detailed computational assessment confirmed that the realignment process introduced minimal inference overhead, ensuring that representational improvements did not compromise model efficiency. The findings reinforced the broader significance of structured representation learning, illustrating that hierarchical embedding modifications could serve as an effective strategy for refining latent space distributions while preserving pre-learned semantic associations. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03766v1-abstract-full').style.display = 'none'; document.getElementById('2502.03766v1-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/2502.03757">arXiv:2502.03757</a> <span> [<a href="https://arxiv.org/pdf/2502.03757">pdf</a>, <a href="https://arxiv.org/ps/2502.03757">ps</a>, <a href="https://arxiv.org/format/2502.03757">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Symbolic Computation">cs.SC</span> </div> </div> <p class="title is-5 mathjax"> Non-minimality of minimal telescopers explained by residues </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+S">Shaoshi Chen</a>, <a href="/search/cs?searchtype=author&query=Kauers%2C+M">Manuel Kauers</a>, <a href="/search/cs?searchtype=author&query=Koutschan%2C+C">Christoph Koutschan</a>, <a href="/search/cs?searchtype=author&query=Li%2C+X">Xiuyun Li</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rong-Hua Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yisen 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="2502.03757v2-abstract-short" style="display: inline;"> Elaborating on an approach recently proposed by Mark van Hoeij, we continue to investigate why creative telescoping occasionally fails to find the minimal-order annihilating operator of a given definite sum or integral. We offer an explanation based on the consideration of residues. </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.03757v2-abstract-full" style="display: none;"> Elaborating on an approach recently proposed by Mark van Hoeij, we continue to investigate why creative telescoping occasionally fails to find the minimal-order annihilating operator of a given definite sum or integral. We offer an explanation based on the consideration of residues. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03757v2-abstract-full').style.display = 'none'; document.getElementById('2502.03757v2-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">v1</span> submitted 5 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">20 pages</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">MSC Class:</span> 33F10 <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.1.2 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.03703">arXiv:2502.03703</a> <span> [<a href="https://arxiv.org/pdf/2502.03703">pdf</a>, <a href="https://arxiv.org/format/2502.03703">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"> On the Expressive Power of Subgraph Graph Neural Networks for Graphs with Bounded Cycles </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chen%2C+Z">Ziang Chen</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Q">Qiao Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzhong 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="2502.03703v1-abstract-short" style="display: inline;"> Graph neural networks (GNNs) have been widely used in graph-related contexts. It is known that the separation power of GNNs is equivalent to that of the Weisfeiler-Lehman (WL) test; hence, GNNs are imperfect at identifying all non-isomorphic graphs, which severely limits their expressive power. This work investigates $k$-hop subgraph GNNs that aggregate information from neighbors with distances up… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03703v1-abstract-full').style.display = 'inline'; document.getElementById('2502.03703v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.03703v1-abstract-full" style="display: none;"> Graph neural networks (GNNs) have been widely used in graph-related contexts. It is known that the separation power of GNNs is equivalent to that of the Weisfeiler-Lehman (WL) test; hence, GNNs are imperfect at identifying all non-isomorphic graphs, which severely limits their expressive power. This work investigates $k$-hop subgraph GNNs that aggregate information from neighbors with distances up to $k$ and incorporate the subgraph structure. We prove that under appropriate assumptions, the $k$-hop subgraph GNNs can approximate any permutation-invariant/equivariant continuous function over graphs without cycles of length greater than $2k+1$ within any error tolerance. We also provide an extension to $k$-hop GNNs without incorporating the subgraph structure. Our numerical experiments on established benchmarks and novel architectures validate our theory on the relationship between the information aggregation distance and the cycle size. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.03703v1-abstract-full').style.display = 'none'; document.getElementById('2502.03703v1-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/2502.02941">arXiv:2502.02941</a> <span> [<a href="https://arxiv.org/pdf/2502.02941">pdf</a>, <a href="https://arxiv.org/format/2502.02941">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"> Fast T2T: Optimization Consistency Speeds Up Diffusion-Based Training-to-Testing Solving for Combinatorial Optimization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+Y">Yang Li</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+J">Jinpei Guo</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Runzhong Wang</a>, <a href="/search/cs?searchtype=author&query=Zha%2C+H">Hongyuan Zha</a>, <a href="/search/cs?searchtype=author&query=Yan%2C+J">Junchi Yan</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="2502.02941v1-abstract-short" style="display: inline;"> Diffusion models have recently advanced Combinatorial Optimization (CO) as a powerful backbone for neural solvers. However, their iterative sampling process requiring denoising across multiple noise levels incurs substantial overhead. We propose to learn direct mappings from different noise levels to the optimal solution for a given instance, facilitating high-quality generation with minimal shots… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02941v1-abstract-full').style.display = 'inline'; document.getElementById('2502.02941v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.02941v1-abstract-full" style="display: none;"> Diffusion models have recently advanced Combinatorial Optimization (CO) as a powerful backbone for neural solvers. However, their iterative sampling process requiring denoising across multiple noise levels incurs substantial overhead. We propose to learn direct mappings from different noise levels to the optimal solution for a given instance, facilitating high-quality generation with minimal shots. This is achieved through an optimization consistency training protocol, which, for a given instance, minimizes the difference among samples originating from varying generative trajectories and time steps relative to the optimal solution. The proposed model enables fast single-step solution generation while retaining the option of multi-step sampling to trade for sampling quality, which offers a more effective and efficient alternative backbone for neural solvers. In addition, within the training-to-testing (T2T) framework, to bridge the gap between training on historical instances and solving new instances, we introduce a novel consistency-based gradient search scheme during the test stage, enabling more effective exploration of the solution space learned during training. It is achieved by updating the latent solution probabilities under objective gradient guidance during the alternation of noise injection and denoising steps. We refer to this model as Fast T2T. Extensive experiments on two popular tasks, the Traveling Salesman Problem (TSP) and Maximal Independent Set (MIS), demonstrate the superiority of Fast T2T regarding both solution quality and efficiency, even outperforming LKH given limited time budgets. Notably, Fast T2T with merely one-step generation and one-step gradient search can mostly outperform the SOTA diffusion-based counterparts that require hundreds of steps, while achieving tens of times speedup. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02941v1-abstract-full').style.display = 'none'; document.getElementById('2502.02941v1-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> <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">Published at NeurIPS 2024, the implementation code is available at https://github.com/Thinklab-SJTU/Fast-T2T</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.02205">arXiv:2502.02205</a> <span> [<a href="https://arxiv.org/pdf/2502.02205">pdf</a>, <a href="https://arxiv.org/format/2502.02205">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"> From Uncertain to Safe: Conformal Fine-Tuning of Diffusion Models for Safe PDE Control </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Hu%2C+P">Peiyan Hu</a>, <a href="/search/cs?searchtype=author&query=Qian%2C+X">Xiaowei Qian</a>, <a href="/search/cs?searchtype=author&query=Deng%2C+W">Wenhao Deng</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui Wang</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+H">Haodong Feng</a>, <a href="/search/cs?searchtype=author&query=Feng%2C+R">Ruiqi Feng</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+T">Tao Zhang</a>, <a href="/search/cs?searchtype=author&query=Wei%2C+L">Long Wei</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yue Wang</a>, <a href="/search/cs?searchtype=author&query=Ma%2C+Z">Zhi-Ming Ma</a>, <a href="/search/cs?searchtype=author&query=Wu%2C+T">Tailin 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="2502.02205v1-abstract-short" style="display: inline;"> The application of deep learning for partial differential equation (PDE)-constrained control is gaining increasing attention. However, existing methods rarely consider safety requirements crucial in real-world applications. To address this limitation, we propose Safe Diffusion Models for PDE Control (SafeDiffCon), which introduce the uncertainty quantile as model uncertainty quantification to achi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02205v1-abstract-full').style.display = 'inline'; document.getElementById('2502.02205v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.02205v1-abstract-full" style="display: none;"> The application of deep learning for partial differential equation (PDE)-constrained control is gaining increasing attention. However, existing methods rarely consider safety requirements crucial in real-world applications. To address this limitation, we propose Safe Diffusion Models for PDE Control (SafeDiffCon), which introduce the uncertainty quantile as model uncertainty quantification to achieve optimal control under safety constraints through both post-training and inference phases. Firstly, our approach post-trains a pre-trained diffusion model to generate control sequences that better satisfy safety constraints while achieving improved control objectives via a reweighted diffusion loss, which incorporates the uncertainty quantile estimated using conformal prediction. Secondly, during inference, the diffusion model dynamically adjusts both its generation process and parameters through iterative guidance and fine-tuning, conditioned on control targets while simultaneously integrating the estimated uncertainty quantile. We evaluate SafeDiffCon on three control tasks: 1D Burgers' equation, 2D incompressible fluid, and controlled nuclear fusion problem. Results demonstrate that SafeDiffCon is the only method that satisfies all safety constraints, whereas other classical and deep learning baselines fail. Furthermore, while adhering to safety constraints, SafeDiffCon achieves the best control performance. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02205v1-abstract-full').style.display = 'none'; document.getElementById('2502.02205v1-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> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2502.02195">arXiv:2502.02195</a> <span> [<a href="https://arxiv.org/pdf/2502.02195">pdf</a>, <a href="https://arxiv.org/format/2502.02195">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Geophysics">physics.geo-ph</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"> EFKAN: A KAN-Integrated Neural Operator For Efficient Magnetotelluric Forward Modeling </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+F">Feng Wang</a>, <a href="/search/cs?searchtype=author&query=Qiu%2C+H">Hong Qiu</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yingying Huang</a>, <a href="/search/cs?searchtype=author&query=Gu%2C+X">Xiaozhe Gu</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Renfang Wang</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+B">Bo Yang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2502.02195v1-abstract-short" style="display: inline;"> Magnetotelluric (MT) forward modeling is fundamental for improving the accuracy and efficiency of MT inversion. Neural operators (NOs) have been effectively used for rapid MT forward modeling, demonstrating their promising performance in solving the MT forward modeling-related partial differential equations (PDEs). Particularly, they can obtain the electromagnetic field at arbitrary locations and… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02195v1-abstract-full').style.display = 'inline'; document.getElementById('2502.02195v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2502.02195v1-abstract-full" style="display: none;"> Magnetotelluric (MT) forward modeling is fundamental for improving the accuracy and efficiency of MT inversion. Neural operators (NOs) have been effectively used for rapid MT forward modeling, demonstrating their promising performance in solving the MT forward modeling-related partial differential equations (PDEs). Particularly, they can obtain the electromagnetic field at arbitrary locations and frequencies. In these NOs, the projection layers have been dominated by multi-layer perceptrons (MLPs), which may potentially reduce the accuracy of solution due to they usually suffer from the disadvantages of MLPs, such as lack of interpretability, overfitting, and so on. Therefore, to improve the accuracy of MT forward modeling with NOs and explore the potential alternatives to MLPs, we propose a novel neural operator by extending the Fourier neural operator (FNO) with Kolmogorov-Arnold network (EFKAN). Within the EFKAN framework, the FNO serves as the branch network to calculate the apparent resistivity and phase from the resistivity model in the frequency domain. Meanwhile, the KAN acts as the trunk network to project the resistivity and phase, determined by the FNO, to the desired locations and frequencies. Experimental results demonstrate that the proposed method not only achieves higher accuracy in obtaining apparent resistivity and phase compared to the NO equipped with MLPs at the desired frequencies and locations but also outperforms traditional numerical methods in terms of computational speed. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2502.02195v1-abstract-full').style.display = 'none'; document.getElementById('2502.02195v1-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> <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 Computers & Geosciences</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.19403">arXiv:2501.19403</a> <span> [<a href="https://arxiv.org/pdf/2501.19403">pdf</a>, <a href="https://arxiv.org/format/2501.19403">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"> Redefining Machine Unlearning: A Conformal Prediction-Motivated Approach </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shi%2C+Y">Yingdan Shi</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ren 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="2501.19403v1-abstract-short" style="display: inline;"> Machine unlearning seeks to systematically remove specified data from a trained model, effectively achieving a state as though the data had never been encountered during training. While metrics such as Unlearning Accuracy (UA) and Membership Inference Attack (MIA) provide a baseline for assessing unlearning performance, they fall short of evaluating the completeness and reliability of forgetting.… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19403v1-abstract-full').style.display = 'inline'; document.getElementById('2501.19403v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.19403v1-abstract-full" style="display: none;"> Machine unlearning seeks to systematically remove specified data from a trained model, effectively achieving a state as though the data had never been encountered during training. While metrics such as Unlearning Accuracy (UA) and Membership Inference Attack (MIA) provide a baseline for assessing unlearning performance, they fall short of evaluating the completeness and reliability of forgetting. This is because the ground truth labels remain potential candidates within the scope of uncertainty quantification, leaving gaps in the evaluation of true forgetting. In this paper, we identify critical limitations in existing unlearning metrics and propose enhanced evaluation metrics inspired by conformal prediction. Our metrics can effectively capture the extent to which ground truth labels are excluded from the prediction set. Furthermore, we observe that many existing machine unlearning methods do not achieve satisfactory forgetting performance when evaluated with our new metrics. To address this, we propose an unlearning framework that integrates conformal prediction insights into Carlini & Wagner adversarial attack loss. Extensive experiments on the image classification task demonstrate that our enhanced metrics offer deeper insights into unlearning effectiveness, and that our unlearning framework significantly improves the forgetting quality of unlearning methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19403v1-abstract-full').style.display = 'none'; document.getElementById('2501.19403v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 31 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.19054">arXiv:2501.19054</a> <span> [<a href="https://arxiv.org/pdf/2501.19054">pdf</a>, <a href="https://arxiv.org/format/2501.19054">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="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Text-to-CAD Generation Through Infusing Visual Feedback in Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruiyu Wang</a>, <a href="/search/cs?searchtype=author&query=Yuan%2C+Y">Yu Yuan</a>, <a href="/search/cs?searchtype=author&query=Sun%2C+S">Shizhao Sun</a>, <a href="/search/cs?searchtype=author&query=Bian%2C+J">Jiang Bian</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="2501.19054v2-abstract-short" style="display: inline;"> Creating Computer-Aided Design (CAD) models requires significant expertise and effort. Text-to-CAD, which converts textual descriptions into CAD parametric sequences, is crucial in streamlining this process. Recent studies have utilized ground-truth parametric sequences, known as sequential signals, as supervision to achieve this goal. However, CAD models are inherently multimodal, comprising para… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19054v2-abstract-full').style.display = 'inline'; document.getElementById('2501.19054v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.19054v2-abstract-full" style="display: none;"> Creating Computer-Aided Design (CAD) models requires significant expertise and effort. Text-to-CAD, which converts textual descriptions into CAD parametric sequences, is crucial in streamlining this process. Recent studies have utilized ground-truth parametric sequences, known as sequential signals, as supervision to achieve this goal. However, CAD models are inherently multimodal, comprising parametric sequences and corresponding rendered visual objects. Besides,the rendering process from parametric sequences to visual objects is many-to-one. Therefore, both sequential and visual signals are critical for effective training. In this work, we introduce CADFusion, a framework that uses Large Language Models (LLMs) as the backbone and alternates between two training stages: the sequential learning (SL) stage and the visual feedback (VF) stage. In the SL stage, we train LLMs using ground-truth parametric sequences, enabling the generation of logically coherent parametric sequences. In the VF stage, we reward parametric sequences that render into visually preferred objects and penalize those that do not, allowing LLMs to learn how rendered visual objects are perceived and evaluated. These two stages alternate throughout the training, ensuring balanced learning and preserving benefits of both signals. Experiments demonstrate that CADFusion significantly improves performance, both qualitatively and quantitatively. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19054v2-abstract-full').style.display = 'none'; document.getElementById('2501.19054v2-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">v1</span> submitted 31 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.19050">arXiv:2501.19050</a> <span> [<a href="https://arxiv.org/pdf/2501.19050">pdf</a>, <a href="https://arxiv.org/format/2501.19050">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"> Norm-Bounded Low-Rank Adaptation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruigang Wang</a>, <a href="/search/cs?searchtype=author&query=Dvijotham%2C+K">Krishnamurthy Dvijotham</a>, <a href="/search/cs?searchtype=author&query=Manchester%2C+I+R">Ian R. Manchester</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="2501.19050v1-abstract-short" style="display: inline;"> In this work, we propose norm-bounded low-rank adaptation (NB-LoRA) for parameter-efficient fine tuning. We introduce two parameterizations that allow explicit bounds on each singular value of the weight adaptation matrix, which can therefore satisfy any prescribed unitarily invariant norm bound, including the Schatten norms (e.g., nuclear, Frobenius, spectral norm). The proposed parameterizations… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19050v1-abstract-full').style.display = 'inline'; document.getElementById('2501.19050v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.19050v1-abstract-full" style="display: none;"> In this work, we propose norm-bounded low-rank adaptation (NB-LoRA) for parameter-efficient fine tuning. We introduce two parameterizations that allow explicit bounds on each singular value of the weight adaptation matrix, which can therefore satisfy any prescribed unitarily invariant norm bound, including the Schatten norms (e.g., nuclear, Frobenius, spectral norm). The proposed parameterizations are unconstrained and complete, i.e. they cover all matrices satisfying the prescribed rank and norm constraints. Experiments on vision fine-tuning benchmarks show that the proposed approach can achieve good adaptation performance while avoiding model catastrophic forgetting and also substantially improve robustness to a wide range of hyper-parameters, including adaptation rank, learning rate and number of training epochs. We also explore applications in privacy-preserving model merging and low-rank matrix completion. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.19050v1-abstract-full').style.display = 'none'; document.getElementById('2501.19050v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 31 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.18585">arXiv:2501.18585</a> <span> [<a href="https://arxiv.org/pdf/2501.18585">pdf</a>, <a href="https://arxiv.org/format/2501.18585">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"> Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+Y">Yue Wang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+Q">Qiuzhi Liu</a>, <a href="/search/cs?searchtype=author&query=Xu%2C+J">Jiahao Xu</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+T">Tian Liang</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xingyu Chen</a>, <a href="/search/cs?searchtype=author&query=He%2C+Z">Zhiwei He</a>, <a href="/search/cs?searchtype=author&query=Song%2C+L">Linfeng Song</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+D">Dian Yu</a>, <a href="/search/cs?searchtype=author&query=Li%2C+J">Juntao Li</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+Z">Zhuosheng Zhang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Rui Wang</a>, <a href="/search/cs?searchtype=author&query=Tu%2C+Z">Zhaopeng Tu</a>, <a href="/search/cs?searchtype=author&query=Mi%2C+H">Haitao Mi</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+D">Dong 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="2501.18585v2-abstract-short" style="display: inline;"> Large language models (LLMs) such as OpenAI's o1 have demonstrated remarkable abilities in complex reasoning tasks by scaling test-time compute and exhibiting human-like deep thinking. However, we identify a phenomenon we term underthinking, where o1-like LLMs frequently switch between different reasoning thoughts without sufficiently exploring promising paths to reach a correct solution. This beh… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18585v2-abstract-full').style.display = 'inline'; document.getElementById('2501.18585v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.18585v2-abstract-full" style="display: none;"> Large language models (LLMs) such as OpenAI's o1 have demonstrated remarkable abilities in complex reasoning tasks by scaling test-time compute and exhibiting human-like deep thinking. However, we identify a phenomenon we term underthinking, where o1-like LLMs frequently switch between different reasoning thoughts without sufficiently exploring promising paths to reach a correct solution. This behavior leads to inadequate depth of reasoning and decreased performance, particularly on challenging mathematical problems. To systematically analyze this issue, we conduct experiments on three challenging test sets and two representative open-source o1-like models, revealing that frequent thought switching correlates with incorrect responses. We introduce a novel metric to quantify underthinking by measuring token efficiency in incorrect answers. To address underthinking, we propose a decoding strategy with thought switching penalty TIP that discourages premature transitions between thoughts, encouraging deeper exploration of each reasoning path. Experimental results demonstrate that our approach improves accuracy across challenging datasets without requiring model fine-tuning. Our findings contribute to understanding reasoning inefficiencies in o1-like LLMs and offer a practical solution to enhance their problem-solving capabilities. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18585v2-abstract-full').style.display = 'none'; document.getElementById('2501.18585v2-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 30 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 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">1. We have updated the results for DeepSeek-R1, and all of our original conclusions remain valid. 2. Our proposed Tip approach remains effective in Best-of-N scenarios (e.g., self-consistency and Laconic Decoding) when built on DeepSeek-R1</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.18564">arXiv:2501.18564</a> <span> [<a href="https://arxiv.org/pdf/2501.18564">pdf</a>, <a href="https://arxiv.org/format/2501.18564">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"> SAM2Act: Integrating Visual Foundation Model with A Memory Architecture for Robotic Manipulation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Fang%2C+H">Haoquan Fang</a>, <a href="/search/cs?searchtype=author&query=Grotz%2C+M">Markus Grotz</a>, <a href="/search/cs?searchtype=author&query=Pumacay%2C+W">Wilbert Pumacay</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y+R">Yi Ru Wang</a>, <a href="/search/cs?searchtype=author&query=Fox%2C+D">Dieter Fox</a>, <a href="/search/cs?searchtype=author&query=Krishna%2C+R">Ranjay Krishna</a>, <a href="/search/cs?searchtype=author&query=Duan%2C+J">Jiafei Duan</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="2501.18564v2-abstract-short" style="display: inline;"> Robotic manipulation systems operating in diverse, dynamic environments must exhibit three critical abilities: multitask interaction, generalization to unseen scenarios, and spatial memory. While significant progress has been made in robotic manipulation, existing approaches often fall short in generalization to complex environmental variations and addressing memory-dependent tasks. To bridge this… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18564v2-abstract-full').style.display = 'inline'; document.getElementById('2501.18564v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.18564v2-abstract-full" style="display: none;"> Robotic manipulation systems operating in diverse, dynamic environments must exhibit three critical abilities: multitask interaction, generalization to unseen scenarios, and spatial memory. While significant progress has been made in robotic manipulation, existing approaches often fall short in generalization to complex environmental variations and addressing memory-dependent tasks. To bridge this gap, we introduce SAM2Act, a multi-view robotic transformer-based policy that leverages multi-resolution upsampling with visual representations from large-scale foundation model. SAM2Act achieves a state-of-the-art average success rate of 86.8% across 18 tasks in the RLBench benchmark, and demonstrates robust generalization on The Colosseum benchmark, with only a 4.3% performance gap under diverse environmental perturbations. Building on this foundation, we propose SAM2Act+, a memory-based architecture inspired by SAM2, which incorporates a memory bank, an encoder, and an attention mechanism to enhance spatial memory. To address the need for evaluating memory-dependent tasks, we introduce MemoryBench, a novel benchmark designed to assess spatial memory and action recall in robotic manipulation. SAM2Act+ achieves competitive performance on MemoryBench, significantly outperforming existing approaches and pushing the boundaries of memory-enabled robotic systems. Project page: https://sam2act.github.io/ <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18564v2-abstract-full').style.display = 'none'; document.getElementById('2501.18564v2-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 30 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 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">Including Appendix, Project page: https://sam2act.github.io/</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.18060">arXiv:2501.18060</a> <span> [<a href="https://arxiv.org/pdf/2501.18060">pdf</a>, <a href="https://arxiv.org/format/2501.18060">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Methodology">stat.ME</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="Machine Learning">stat.ML</span> </div> </div> <p class="title is-5 mathjax"> Noise-Adaptive Conformal Classification with Marginal Coverage </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Bortolotti%2C+T">Teresa Bortolotti</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+Y+X+R">Y. X. Rachel Wang</a>, <a href="/search/cs?searchtype=author&query=Tong%2C+X">Xin Tong</a>, <a href="/search/cs?searchtype=author&query=Menafoglio%2C+A">Alessandra Menafoglio</a>, <a href="/search/cs?searchtype=author&query=Vantini%2C+S">Simone Vantini</a>, <a href="/search/cs?searchtype=author&query=Sesia%2C+M">Matteo Sesia</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="2501.18060v1-abstract-short" style="display: inline;"> Conformal inference provides a rigorous statistical framework for uncertainty quantification in machine learning, enabling well-calibrated prediction sets with precise coverage guarantees for any classification model. However, its reliance on the idealized assumption of perfect data exchangeability limits its effectiveness in the presence of real-world complications, such as low-quality labels --… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18060v1-abstract-full').style.display = 'inline'; document.getElementById('2501.18060v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.18060v1-abstract-full" style="display: none;"> Conformal inference provides a rigorous statistical framework for uncertainty quantification in machine learning, enabling well-calibrated prediction sets with precise coverage guarantees for any classification model. However, its reliance on the idealized assumption of perfect data exchangeability limits its effectiveness in the presence of real-world complications, such as low-quality labels -- a widespread issue in modern large-scale data sets. This work tackles this open problem by introducing an adaptive conformal inference method capable of efficiently handling deviations from exchangeability caused by random label noise, leading to informative prediction sets with tight marginal coverage guarantees even in those challenging scenarios. We validate our method through extensive numerical experiments demonstrating its effectiveness on synthetic and real data sets, including CIFAR-10H and BigEarthNet. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.18060v1-abstract-full').style.display = 'none'; document.getElementById('2501.18060v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 29 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.17116">arXiv:2501.17116</a> <span> [<a href="https://arxiv.org/pdf/2501.17116">pdf</a>, <a href="https://arxiv.org/format/2501.17116">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> </div> </div> <p class="title is-5 mathjax"> Optimizing Large Language Model Training Using FP4 Quantization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruizhe Wang</a>, <a href="/search/cs?searchtype=author&query=Gong%2C+Y">Yeyun Gong</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+X">Xiao Liu</a>, <a href="/search/cs?searchtype=author&query=Zhao%2C+G">Guoshuai Zhao</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Z">Ziyue Yang</a>, <a href="/search/cs?searchtype=author&query=Guo%2C+B">Baining Guo</a>, <a href="/search/cs?searchtype=author&query=Zha%2C+Z">Zhengjun Zha</a>, <a href="/search/cs?searchtype=author&query=Cheng%2C+P">Peng Cheng</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="2501.17116v1-abstract-short" style="display: inline;"> The growing computational demands of training large language models (LLMs) necessitate more efficient methods. Quantized training presents a promising solution by enabling low-bit arithmetic operations to reduce these costs. While FP8 precision has demonstrated feasibility, leveraging FP4 remains a challenge due to significant quantization errors and limited representational capacity. This work in… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.17116v1-abstract-full').style.display = 'inline'; document.getElementById('2501.17116v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.17116v1-abstract-full" style="display: none;"> The growing computational demands of training large language models (LLMs) necessitate more efficient methods. Quantized training presents a promising solution by enabling low-bit arithmetic operations to reduce these costs. While FP8 precision has demonstrated feasibility, leveraging FP4 remains a challenge due to significant quantization errors and limited representational capacity. This work introduces the first FP4 training framework for LLMs, addressing these challenges with two key innovations: a differentiable quantization estimator for precise weight updates and an outlier clamping and compensation strategy to prevent activation collapse. To ensure stability, the framework integrates a mixed-precision training scheme and vector-wise quantization. Experimental results demonstrate that our FP4 framework achieves accuracy comparable to BF16 and FP8, with minimal degradation, scaling effectively to 13B-parameter LLMs trained on up to 100B tokens. With the emergence of next-generation hardware supporting FP4, our framework sets a foundation for efficient ultra-low precision training. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.17116v1-abstract-full').style.display = 'none'; document.getElementById('2501.17116v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 28 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 2025. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2501.16368">arXiv:2501.16368</a> <span> [<a href="https://arxiv.org/pdf/2501.16368">pdf</a>, <a href="https://arxiv.org/format/2501.16368">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="Systems and Control">eess.SY</span> </div> </div> <p class="title is-5 mathjax"> Foundation Models for CPS-IoT: Opportunities and Challenges </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Baris%2C+O">Ozan Baris</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+Y">Yizhuo Chen</a>, <a href="/search/cs?searchtype=author&query=Dong%2C+G">Gaofeng Dong</a>, <a href="/search/cs?searchtype=author&query=Han%2C+L">Liying Han</a>, <a href="/search/cs?searchtype=author&query=Kimura%2C+T">Tomoyoshi Kimura</a>, <a href="/search/cs?searchtype=author&query=Quan%2C+P">Pengrui Quan</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+R">Ruijie Wang</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+T">Tianchen Wang</a>, <a href="/search/cs?searchtype=author&query=Abdelzaher%2C+T">Tarek Abdelzaher</a>, <a href="/search/cs?searchtype=author&query=Berg%C3%A9s%2C+M">Mario Berg茅s</a>, <a href="/search/cs?searchtype=author&query=Liang%2C+P+P">Paul Pu Liang</a>, <a href="/search/cs?searchtype=author&query=Srivastava%2C+M">Mani Srivastava</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="2501.16368v2-abstract-short" style="display: inline;"> Methods from machine learning (ML) have transformed the implementation of Perception-Cognition-Communication-Action loops in Cyber-Physical Systems (CPS) and the Internet of Things (IoT), replacing mechanistic and basic statistical models with those derived from data. However, the first generation of ML approaches, which depend on supervised learning with annotated data to create task-specific mod… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.16368v2-abstract-full').style.display = 'inline'; document.getElementById('2501.16368v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2501.16368v2-abstract-full" style="display: none;"> Methods from machine learning (ML) have transformed the implementation of Perception-Cognition-Communication-Action loops in Cyber-Physical Systems (CPS) and the Internet of Things (IoT), replacing mechanistic and basic statistical models with those derived from data. However, the first generation of ML approaches, which depend on supervised learning with annotated data to create task-specific models, faces significant limitations in scaling to the diverse sensor modalities, deployment configurations, application tasks, and operating dynamics characterizing real-world CPS-IoT systems. The success of task-agnostic foundation models (FMs), including multimodal large language models (LLMs), in addressing similar challenges across natural language, computer vision, and human speech has generated considerable enthusiasm for and exploration of FMs and LLMs as flexible building blocks in CPS-IoT analytics pipelines, promising to reduce the need for costly task-specific engineering. Nonetheless, a significant gap persists between the current capabilities of FMs and LLMs in the CPS-IoT domain and the requirements they must meet to be viable for CPS-IoT applications. In this paper, we analyze and characterize this gap through a thorough examination of the state of the art and our research, which extends beyond it in various dimensions. Based on the results of our analysis and research, we identify essential desiderata that CPS-IoT domain-specific FMs and LLMs must satisfy to bridge this gap. We also propose actions by CPS-IoT researchers to collaborate in developing key community resources necessary for establishing FMs and LLMs as foundational tools for the next generation of CPS-IoT systems. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2501.16368v2-abstract-full').style.display = 'none'; document.getElementById('2501.16368v2-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">v1</span> submitted 22 January, 2025; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> January 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=Wang%2C+R&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Wang%2C+R&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