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 532 results for author: <span class="mathjax">Kumar, 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=Kumar%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="Kumar, 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=Kumar%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="Kumar, 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=Kumar%2C+R&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%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/2411.13251">arXiv:2411.13251</a> <span> [<a href="https://arxiv.org/pdf/2411.13251">pdf</a>, <a href="https://arxiv.org/format/2411.13251">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="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"> BelHouse3D: A Benchmark Dataset for Assessing Occlusion Robustness in 3D Point Cloud Semantic Segmentation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+U+R">Umamaheswaran Raman Kumar</a>, <a href="/search/cs?searchtype=author&query=Fayjie%2C+A+R">Abdur Razzaq Fayjie</a>, <a href="/search/cs?searchtype=author&query=Hannaert%2C+J">Jurgen Hannaert</a>, <a href="/search/cs?searchtype=author&query=Vandewalle%2C+P">Patrick Vandewalle</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.13251v1-abstract-short" style="display: inline;"> Large-scale 2D datasets have been instrumental in advancing machine learning; however, progress in 3D vision tasks has been relatively slow. This disparity is largely due to the limited availability of 3D benchmarking datasets. In particular, creating real-world point cloud datasets for indoor scene semantic segmentation presents considerable challenges, including data collection within confined s… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.13251v1-abstract-full').style.display = 'inline'; document.getElementById('2411.13251v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.13251v1-abstract-full" style="display: none;"> Large-scale 2D datasets have been instrumental in advancing machine learning; however, progress in 3D vision tasks has been relatively slow. This disparity is largely due to the limited availability of 3D benchmarking datasets. In particular, creating real-world point cloud datasets for indoor scene semantic segmentation presents considerable challenges, including data collection within confined spaces and the costly, often inaccurate process of per-point labeling to generate ground truths. While synthetic datasets address some of these challenges, they often fail to replicate real-world conditions, particularly the occlusions that occur in point clouds collected from real environments. Existing 3D benchmarking datasets typically evaluate deep learning models under the assumption that training and test data are independently and identically distributed (IID), which affects the models' usability for real-world point cloud segmentation. To address these challenges, we introduce the BelHouse3D dataset, a new synthetic point cloud dataset designed for 3D indoor scene semantic segmentation. This dataset is constructed using real-world references from 32 houses in Belgium, ensuring that the synthetic data closely aligns with real-world conditions. Additionally, we include a test set with data occlusion to simulate out-of-distribution (OOD) scenarios, reflecting the occlusions commonly encountered in real-world point clouds. We evaluate popular point-based semantic segmentation methods using our OOD setting and present a benchmark. We believe that BelHouse3D and its OOD setting will advance research in 3D point cloud semantic segmentation for indoor scenes, providing valuable insights for the development of more generalizable models. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.13251v1-abstract-full').style.display = 'none'; document.getElementById('2411.13251v1-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">20 pages, 6 figures, 3 tables, accepted at ECCV 2024 Workshops</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.07501">arXiv:2411.07501</a> <span> [<a href="https://arxiv.org/pdf/2411.07501">pdf</a>, <a href="https://arxiv.org/format/2411.07501">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="Computer Vision and Pattern Recognition">cs.CV</span> </div> </div> <p class="title is-5 mathjax"> LAuReL: Learned Augmented Residual Layer </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Menghani%2C+G">Gaurav Menghani</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S">Sanjiv Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.07501v2-abstract-short" style="display: inline;"> One of the core pillars of efficient deep learning methods is architectural improvements such as the residual/skip connection, which has led to significantly better model convergence and quality. Since then the residual connection has become ubiquitous in not just convolutional neural networks but also transformer-based architectures, the backbone of LLMs. In this paper we introduce \emph{Learne… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.07501v2-abstract-full').style.display = 'inline'; document.getElementById('2411.07501v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.07501v2-abstract-full" style="display: none;"> One of the core pillars of efficient deep learning methods is architectural improvements such as the residual/skip connection, which has led to significantly better model convergence and quality. Since then the residual connection has become ubiquitous in not just convolutional neural networks but also transformer-based architectures, the backbone of LLMs. In this paper we introduce \emph{Learned Augmented Residual Layer} (LAuReL) -- a novel generalization of the canonical residual connection -- with the goal to be an in-situ replacement of the latter while outperforming on both model quality and footprint metrics. Our experiments show that using \laurel can help boost performance for both vision and language models. For example, on the ResNet-50, ImageNet 1K task, it achieves $60\%$ of the gains from adding an extra layer, while only adding $0.003\%$ more parameters, and matches it while adding $2.6\times$ fewer parameters. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.07501v2-abstract-full').style.display = 'none'; document.getElementById('2411.07501v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 13 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 11 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted at the 2nd Efficient Systems for Foundation Models Workshop at the International Conference on Machine Learning (ICML) 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.04843">arXiv:2411.04843</a> <span> [<a href="https://arxiv.org/pdf/2411.04843">pdf</a>, <a href="https://arxiv.org/format/2411.04843">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Science and Game Theory">cs.GT</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Learning in Budgeted Auctions with Spacing Objectives </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Fikioris%2C+G">Giannis Fikioris</a>, <a href="/search/cs?searchtype=author&query=Kleinberg%2C+R">Robert Kleinberg</a>, <a href="/search/cs?searchtype=author&query=Kolumbus%2C+Y">Yoav Kolumbus</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Raunak Kumar</a>, <a href="/search/cs?searchtype=author&query=Mansour%2C+Y">Yishay Mansour</a>, <a href="/search/cs?searchtype=author&query=Tardos%2C+%C3%89">脡va Tardos</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.04843v1-abstract-short" style="display: inline;"> In many repeated auction settings, participants care not only about how frequently they win but also how their winnings are distributed over time. This problem arises in various practical domains where avoiding congested demand is crucial, such as online retail sales and compute services, as well as in advertising campaigns that require sustained visibility over time. We introduce a simple model o… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.04843v1-abstract-full').style.display = 'inline'; document.getElementById('2411.04843v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.04843v1-abstract-full" style="display: none;"> In many repeated auction settings, participants care not only about how frequently they win but also how their winnings are distributed over time. This problem arises in various practical domains where avoiding congested demand is crucial, such as online retail sales and compute services, as well as in advertising campaigns that require sustained visibility over time. We introduce a simple model of this phenomenon, modeling it as a budgeted auction where the value of a win is a concave function of the time since the last win. This implies that for a given number of wins, even spacing over time is optimal. We also extend our model and results to the case when not all wins result in "conversions" (realization of actual gains), and the probability of conversion depends on a context. The goal is to maximize and evenly space conversions rather than just wins. We study the optimal policies for this setting in second-price auctions and offer learning algorithms for the bidders that achieve low regret against the optimal bidding policy in a Bayesian online setting. Our main result is a computationally efficient online learning algorithm that achieves $\tilde O(\sqrt T)$ regret. We achieve this by showing that an infinite-horizon Markov decision process (MDP) with the budget constraint in expectation is essentially equivalent to our problem, even when limiting that MDP to a very small number of states. The algorithm achieves low regret by learning a bidding policy that chooses bids as a function of the context and the system's state, which will be the time elapsed since the last win (or conversion). We show that state-independent strategies incur linear regret even without uncertainty of conversions. We complement this by showing that there are state-independent strategies that, while still having linear regret, achieve a $(1-\frac 1 e)$ approximation to the optimal reward. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.04843v1-abstract-full').style.display = 'none'; document.getElementById('2411.04843v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 7 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.04205">arXiv:2411.04205</a> <span> [<a href="https://arxiv.org/pdf/2411.04205">pdf</a>, <a href="https://arxiv.org/format/2411.04205">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="Cryptography and Security">cs.CR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Data Structures and Algorithms">cs.DS</span> </div> </div> <p class="title is-5 mathjax"> Scalable DP-SGD: Shuffling vs. Poisson Subsampling </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chua%2C+L">Lynn Chua</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Amer Sinha</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan Zhang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.04205v1-abstract-short" style="display: inline;"> We provide new lower bounds on the privacy guarantee of the multi-epoch Adaptive Batch Linear Queries (ABLQ) mechanism with shuffled batch sampling, demonstrating substantial gaps when compared to Poisson subsampling; prior analysis was limited to a single epoch. Since the privacy analysis of Differentially Private Stochastic Gradient Descent (DP-SGD) is obtained by analyzing the ABLQ mechanism, t… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.04205v1-abstract-full').style.display = 'inline'; document.getElementById('2411.04205v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.04205v1-abstract-full" style="display: none;"> We provide new lower bounds on the privacy guarantee of the multi-epoch Adaptive Batch Linear Queries (ABLQ) mechanism with shuffled batch sampling, demonstrating substantial gaps when compared to Poisson subsampling; prior analysis was limited to a single epoch. Since the privacy analysis of Differentially Private Stochastic Gradient Descent (DP-SGD) is obtained by analyzing the ABLQ mechanism, this brings into serious question the common practice of implementing shuffling-based DP-SGD, but reporting privacy parameters as if Poisson subsampling was used. To understand the impact of this gap on the utility of trained machine learning models, we introduce a practical approach to implement Poisson subsampling at scale using massively parallel computation, and efficiently train models with the same. We compare the utility of models trained with Poisson-subsampling-based DP-SGD, and the optimistic estimates of utility when using shuffling, via our new lower bounds on the privacy guarantee of ABLQ with shuffling. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.04205v1-abstract-full').style.display = 'none'; document.getElementById('2411.04205v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 6 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">To appear at NeurIPS 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.03581">arXiv:2411.03581</a> <span> [<a href="https://arxiv.org/pdf/2411.03581">pdf</a>, <a href="https://arxiv.org/format/2411.03581">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Systems and Control">eess.SY</span> </div> </div> <p class="title is-5 mathjax"> Can Robotic Cues Manipulate Human Decisions? Exploring Consensus Building via Bias-Controlled Non-linear Opinion Dynamics and Robotic Eye Gaze Mediated Interaction in Human-Robot Teaming </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rajul Kumar</a>, <a href="/search/cs?searchtype=author&query=Bhatti%2C+A">Adam Bhatti</a>, <a href="/search/cs?searchtype=author&query=Yao%2C+N">Ningshi Yao</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.03581v1-abstract-short" style="display: inline;"> Although robots are becoming more advanced with human-like anthropomorphic features and decision-making abilities to improve collaboration, the active integration of humans into this process remains under-explored. This article presents the first experimental study exploring decision-making interactions between humans and robots with visual cues from robotic eyes, which can dynamically influence h… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.03581v1-abstract-full').style.display = 'inline'; document.getElementById('2411.03581v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.03581v1-abstract-full" style="display: none;"> Although robots are becoming more advanced with human-like anthropomorphic features and decision-making abilities to improve collaboration, the active integration of humans into this process remains under-explored. This article presents the first experimental study exploring decision-making interactions between humans and robots with visual cues from robotic eyes, which can dynamically influence human opinion formation. The cues generated by robotic eyes gradually guide human decisions towards alignment with the robot's choices. Both human and robot decision-making processes are modeled as non-linear opinion dynamics with evolving biases. To examine these opinion dynamics under varying biases, we conduct numerical parametric and equilibrium continuation analyses using tuned parameters designed explicitly for the presented human-robot interaction experiment. Furthermore, to facilitate the transition from disagreement to agreement, we introduced a human opinion observation algorithm integrated with the formation of the robot's opinion, where the robot's behavior is controlled based on its formed opinion. The algorithms developed aim to enhance human involvement in consensus building, fostering effective collaboration between humans and robots. Experiments with 51 participants (N = 51) show that human-robot teamwork can be improved by guiding human decisions using robotic cues. Finally, we provide detailed insights on the effects of trust, cognitive load, and participant demographics on decision-making based on user feedback and post-experiment interviews. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.03581v1-abstract-full').style.display = 'none'; document.getElementById('2411.03581v1-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">35 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/2411.02833">arXiv:2411.02833</a> <span> [<a href="https://arxiv.org/pdf/2411.02833">pdf</a>, <a href="https://arxiv.org/format/2411.02833">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 class="is-inline-block" style="margin-left: 0.5rem"> <div class="tags has-addons"> <span class="tag is-dark is-size-7">doi</span> <span class="tag is-light is-size-7"><a class="" href="https://doi.org/10.1145/3702250.3702254">10.1145/3702250.3702254 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> Lost in Context: The Influence of Context on Feature Attribution Methods for Object Recognition </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Adhikari%2C+S">Sayanta Adhikari</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rishav Kumar</a>, <a href="/search/cs?searchtype=author&query=Mopuri%2C+K+R">Konda Reddy Mopuri</a>, <a href="/search/cs?searchtype=author&query=Pachamuthu%2C+R">Rajalakshmi Pachamuthu</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.02833v1-abstract-short" style="display: inline;"> Contextual information plays a critical role in object recognition models within computer vision, where changes in context can significantly affect accuracy, underscoring models' dependence on contextual cues. This study investigates how context manipulation influences both model accuracy and feature attribution, providing insights into the reliance of object recognition models on contextual infor… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.02833v1-abstract-full').style.display = 'inline'; document.getElementById('2411.02833v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.02833v1-abstract-full" style="display: none;"> Contextual information plays a critical role in object recognition models within computer vision, where changes in context can significantly affect accuracy, underscoring models' dependence on contextual cues. This study investigates how context manipulation influences both model accuracy and feature attribution, providing insights into the reliance of object recognition models on contextual information as understood through the lens of feature attribution methods. We employ a range of feature attribution techniques to decipher the reliance of deep neural networks on context in object recognition tasks. Using the ImageNet-9 and our curated ImageNet-CS datasets, we conduct experiments to evaluate the impact of contextual variations, analyzed through feature attribution methods. Our findings reveal several key insights: (a) Correctly classified images predominantly emphasize object volume attribution over context volume attribution. (b) The dependence on context remains relatively stable across different context modifications, irrespective of classification accuracy. (c) Context change exerts a more pronounced effect on model performance than Context perturbations. (d) Surprisingly, context attribution in `no-information' scenarios is non-trivial. Our research moves beyond traditional methods by assessing the implications of broad-level modifications on object recognition, either in the object or its context. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.02833v1-abstract-full').style.display = 'none'; document.getElementById('2411.02833v1-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 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Published in ICVGIP 2024</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> I.4.m; I.2.10 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2411.00643">arXiv:2411.00643</a> <span> [<a href="https://arxiv.org/pdf/2411.00643">pdf</a>, <a href="https://arxiv.org/format/2411.00643">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"> Transforming Agriculture: Exploring Diverse Practices and Technological Innovations </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ramakant Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2411.00643v1-abstract-short" style="display: inline;"> Agriculture is a vital sector that significantly contributes to the economy and food security, particularly in regions like Varanasi, India. This paper explores various types of agriculture practiced in the area, including subsistence, commercial, intensive, extensive, industrial, organic, agroforestry, aquaculture, and urban agriculture. Each type presents unique challenges and opportunities, nec… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.00643v1-abstract-full').style.display = 'inline'; document.getElementById('2411.00643v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2411.00643v1-abstract-full" style="display: none;"> Agriculture is a vital sector that significantly contributes to the economy and food security, particularly in regions like Varanasi, India. This paper explores various types of agriculture practiced in the area, including subsistence, commercial, intensive, extensive, industrial, organic, agroforestry, aquaculture, and urban agriculture. Each type presents unique challenges and opportunities, necessitating innovative approaches to enhance productivity and sustainability. To address these challenges, the integration of advanced technologies such as sensors and communication protocols is essential. Sensors can provide real-time data on soil health, moisture levels, and crop conditions, enabling farmers to make informed decisions. Communication technologies facilitate the seamless transfer of this data, allowing for timely interventions and optimized resource management. Moreover, programming techniques play a crucial role in developing applications that process and analyze agricultural data. By leveraging machine learning algorithms, farmers can gain insights into crop performance, predict yields, and implement precision agriculture practices. This paper highlights the significance of combining traditional agricultural practices with modern technologies to create a resilient agricultural ecosystem. The findings underscore the potential of integrating sensors, communication technologies, and programming in transforming agricultural practices in Varanasi. By fostering a data-driven approach, this research aims to contribute to sustainable farming, enhance food security, and improve the livelihoods of farmers in the region. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2411.00643v1-abstract-full').style.display = 'none'; document.getElementById('2411.00643v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 1 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> November 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.23123">arXiv:2410.23123</a> <span> [<a href="https://arxiv.org/pdf/2410.23123">pdf</a>, <a href="https://arxiv.org/format/2410.23123">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"> On Memorization of Large Language Models in Logical Reasoning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Xie%2C+C">Chulin Xie</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yangsibo Huang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan Zhang</a>, <a href="/search/cs?searchtype=author&query=Yu%2C+D">Da Yu</a>, <a href="/search/cs?searchtype=author&query=Chen%2C+X">Xinyun Chen</a>, <a href="/search/cs?searchtype=author&query=Lin%2C+B+Y">Bill Yuchen Lin</a>, <a href="/search/cs?searchtype=author&query=Li%2C+B">Bo Li</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.23123v1-abstract-short" style="display: inline;"> Large language models (LLMs) achieve good performance on challenging reasoning benchmarks, yet could also make basic reasoning mistakes. This contrasting behavior is puzzling when it comes to understanding the mechanisms behind LLMs' reasoning capabilities. One hypothesis is that the increasingly high and nearly saturated performance on common reasoning benchmarks could be due to the memorization… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.23123v1-abstract-full').style.display = 'inline'; document.getElementById('2410.23123v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.23123v1-abstract-full" style="display: none;"> Large language models (LLMs) achieve good performance on challenging reasoning benchmarks, yet could also make basic reasoning mistakes. This contrasting behavior is puzzling when it comes to understanding the mechanisms behind LLMs' reasoning capabilities. One hypothesis is that the increasingly high and nearly saturated performance on common reasoning benchmarks could be due to the memorization of similar problems. In this paper, we systematically investigate this hypothesis with a quantitative measurement of memorization in reasoning tasks, using a dynamically generated logical reasoning benchmark based on Knights and Knaves (K&K) puzzles. We found that LLMs could interpolate the training puzzles (achieving near-perfect accuracy) after fine-tuning, yet fail when those puzzles are slightly perturbed, suggesting that the models heavily rely on memorization to solve those training puzzles. On the other hand, we show that while fine-tuning leads to heavy memorization, it also consistently improves generalization performance. In-depth analyses with perturbation tests, cross difficulty-level transferability, probing model internals, and fine-tuning with wrong answers suggest that the LLMs learn to reason on K&K puzzles despite training data memorization. This phenomenon indicates that LLMs exhibit a complex interplay between memorization and genuine reasoning abilities. Finally, our analysis with per-sample memorization score sheds light on how LLMs switch between reasoning and memorization in solving logical puzzles. Our code and data are available at https://memkklogic.github.io. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.23123v1-abstract-full').style.display = 'none'; document.getElementById('2410.23123v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 30 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.20484">arXiv:2410.20484</a> <span> [<a href="https://arxiv.org/pdf/2410.20484">pdf</a>, <a href="https://arxiv.org/ps/2410.20484">ps</a>, <a href="https://arxiv.org/format/2410.20484">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Emerging Technologies">cs.ET</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Distributed, Parallel, and Cluster Computing">cs.DC</span> </div> </div> <p class="title is-5 mathjax"> Smart Space Environments: Key Challenges and Innovative Solutions </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ramakant Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.20484v1-abstract-short" style="display: inline;"> The integration of LoRaWAN (Long Range Wide Area Network) technology with both active and passive sensors presents a transformative opportunity for the development of smart home systems. This paper explores how active sensors, such as motion detectors and ultrasonic sensors, and passive sensors, including temperature and humidity sensors, work together to enhance connectivity and efficiency within… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20484v1-abstract-full').style.display = 'inline'; document.getElementById('2410.20484v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.20484v1-abstract-full" style="display: none;"> The integration of LoRaWAN (Long Range Wide Area Network) technology with both active and passive sensors presents a transformative opportunity for the development of smart home systems. This paper explores how active sensors, such as motion detectors and ultrasonic sensors, and passive sensors, including temperature and humidity sensors, work together to enhance connectivity and efficiency within diverse environments while addressing the challenges of modern living. By leveraging LoRaWAN long-range capabilities and low power consumption, the proposed framework enables effective data transmission from remote sensors, facilitating applications such as smart agriculture, environmental monitoring, and comprehensive home automation. Active sensors emit energy to detect changes in their surroundings, providing real-time data crucial for security and automation, while passive sensors capture ambient energy to monitor environmental conditions, ensuring resource efficiency and user comfort. The synergy between LoRaWAN and these various sensor types promotes innovation, contributing to a more responsive and sustainable living experience. Furthermore, this research highlights the adaptability of the proposed system, allowing for seamless integration of new devices and advanced functionalities. As the landscape of smart home technology continues to evolve, ongoing research in this area will yield advanced solutions tailored to user needs, ultimately paving the way for smarter, safer, and more efficient living environments. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20484v1-abstract-full').style.display = 'none'; document.getElementById('2410.20484v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 27 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.20231">arXiv:2410.20231</a> <span> [<a href="https://arxiv.org/pdf/2410.20231">pdf</a>, <a href="https://arxiv.org/format/2410.20231">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"> CAVE: Classifying Abnormalities in Video Capsule Endoscopy </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Harish%2C+I">Ishita Harish</a>, <a href="/search/cs?searchtype=author&query=Mishra%2C+S">Saurav Mishra</a>, <a href="/search/cs?searchtype=author&query=Bhadoria%2C+N">Neha Bhadoria</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rithik Kumar</a>, <a href="/search/cs?searchtype=author&query=Arora%2C+M">Madhav Arora</a>, <a href="/search/cs?searchtype=author&query=Zahra%2C+S+R">Syed Rameem Zahra</a>, <a href="/search/cs?searchtype=author&query=Gupta%2C+A">Ankur Gupta</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.20231v1-abstract-short" style="display: inline;"> In this study, we explore an ensemble-based approach to improve classification accuracy in complex image datasets. Utilizing a Convolutional Block Attention Module (CBAM) alongside a Deep Neural Network (DNN) we leverage the unique feature-extraction capabilities of each model to enhance the overall accuracy. Additional models, such as Random Forest, XGBoost, Support Vector Machine (SVM), and K-Ne… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20231v1-abstract-full').style.display = 'inline'; document.getElementById('2410.20231v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.20231v1-abstract-full" style="display: none;"> In this study, we explore an ensemble-based approach to improve classification accuracy in complex image datasets. Utilizing a Convolutional Block Attention Module (CBAM) alongside a Deep Neural Network (DNN) we leverage the unique feature-extraction capabilities of each model to enhance the overall accuracy. Additional models, such as Random Forest, XGBoost, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN), are introduced to further diversify the predictive power of our ensemble. By leveraging these methods, the proposed approach provides robust feature discrimination and improved classification results. Experimental evaluations demonstrate that the ensemble achieves higher accuracy and robustness across challenging and imbalanced classes, showing significant promise for broader applications in computer vision tasks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.20231v1-abstract-full').style.display = 'none'; document.getElementById('2410.20231v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 26 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.19708">arXiv:2410.19708</a> <span> [<a href="https://arxiv.org/pdf/2410.19708">pdf</a>, <a href="https://arxiv.org/ps/2410.19708">ps</a>, <a href="https://arxiv.org/format/2410.19708">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="Emerging Technologies">cs.ET</span> </div> </div> <p class="title is-5 mathjax"> Integrating LoRaWAN with Mobile Ad-hoc Networks for Enhanced Campus Communication </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ramakant Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.19708v1-abstract-short" style="display: inline;"> The integration of Long Range Wide Area Network (LoRaWAN) with Mobile Ad-hoc Networks (MANETs) presents a promising solution for enhancing communication networks within campus environments. This paper explores the unique advantages of combining these two technologies, including scalability, energy efficiency, flexibility, and support for diverse applications. LoRaWAN low power consumption and exte… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.19708v1-abstract-full').style.display = 'inline'; document.getElementById('2410.19708v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.19708v1-abstract-full" style="display: none;"> The integration of Long Range Wide Area Network (LoRaWAN) with Mobile Ad-hoc Networks (MANETs) presents a promising solution for enhancing communication networks within campus environments. This paper explores the unique advantages of combining these two technologies, including scalability, energy efficiency, flexibility, and support for diverse applications. LoRaWAN low power consumption and extended range capabilities address the challenges of traditional communication methods, enabling reliable data transmission across various campus scenarios, such as emergency alerts, event coordination, and real-time monitoring. We also identify key challenges faced in this integrated architecture, including signal interference, data packet collisions, and energy management. By providing a comprehensive survey of existing techniques and solutions categorized by the network protocol stack layers, this study aims to inform future research and development efforts in creating robust, energy efficient communication systems tailored for modern educational institutions. Ultimately, the findings highlight the potential of LoRaWAN MANET architectures to transform campus communication into a more reliable, adaptable, and cost effective framework. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.19708v1-abstract-full').style.display = 'none'; document.getElementById('2410.19708v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.15736">arXiv:2410.15736</a> <span> [<a href="https://arxiv.org/pdf/2410.15736">pdf</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Hardware Architecture">cs.AR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Emerging Technologies">cs.ET</span> </div> </div> <p class="title is-5 mathjax"> Design of a 64-bit SQRT-CSLA with Reduced Area and High-Speed Applications in Low Power VLSI Circuits </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Pallavi%2C+C">CH. Pallavi</a>, <a href="/search/cs?searchtype=author&query=Padma%2C+C">C. Padma</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R+K">R. Kiran Kumar</a>, <a href="/search/cs?searchtype=author&query=Suguna%2C+T">T. Suguna</a>, <a href="/search/cs?searchtype=author&query=Nalini%2C+C">C. Nalini</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.15736v1-abstract-short" style="display: inline;"> The main areas of research in VLSI system design include area, high speed, and power-efficient data route logic systems. The amount of time needed to send a carry through the adder limits the pace at which addition can occur in digital adders. One of the quickest adders, the Carry Select Adder (CSLA), is utilized by various data processing processors to carry out quick arithmetic operations. It is… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.15736v1-abstract-full').style.display = 'inline'; document.getElementById('2410.15736v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.15736v1-abstract-full" style="display: none;"> The main areas of research in VLSI system design include area, high speed, and power-efficient data route logic systems. The amount of time needed to send a carry through the adder limits the pace at which addition can occur in digital adders. One of the quickest adders, the Carry Select Adder (CSLA), is utilized by various data processing processors to carry out quick arithmetic operations. It is evident from the CSLA's structure that there is room to cut back on both the area and the delay. This work employs a straightforward and effective gate-level adjustment (in a regular structure) that significantly lowers the CSLA's area and delay. In light of this adjustment Square-Root Carry Select Adder (SQRT CSLA) designs with bit lengths of 8, 16, 32, and 64. When compared to the standard SQRT CSLA, the suggested design significantly reduces both area and latency. Xilinx ISE tool is used for Simulation and synthesis. The performance of the recommended designs in terms of delay is estimated in this study using the standard designs. The study of the findings indicates that the suggested CSLA structure outperforms the standard SQRT CSLA. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.15736v1-abstract-full').style.display = 'none'; document.getElementById('2410.15736v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 21 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.14709">arXiv:2410.14709</a> <span> [<a href="https://arxiv.org/pdf/2410.14709">pdf</a>, <a href="https://arxiv.org/format/2410.14709">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="Sound">cs.SD</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</span> </div> </div> <p class="title is-5 mathjax"> A two-stage transliteration approach to improve performance of a multilingual ASR </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rohit Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.14709v1-abstract-short" style="display: inline;"> End-to-end Automatic Speech Recognition (ASR) systems are rapidly claiming to become state-of-art over other modeling methods. Several techniques have been introduced to improve their ability to handle multiple languages. However, due to variation in writing scripts for different languages, while decoding acoustically similar units, they do not always map to an appropriate grapheme in the target l… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.14709v1-abstract-full').style.display = 'inline'; document.getElementById('2410.14709v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.14709v1-abstract-full" style="display: none;"> End-to-end Automatic Speech Recognition (ASR) systems are rapidly claiming to become state-of-art over other modeling methods. Several techniques have been introduced to improve their ability to handle multiple languages. However, due to variation in writing scripts for different languages, while decoding acoustically similar units, they do not always map to an appropriate grapheme in the target language. This restricts the scalability and adaptability of the model while dealing with multiple languages in code-mixing scenarios. This paper presents an approach to build a language-agnostic end-to-end model trained on a grapheme set obtained by projecting the multilingual grapheme data to the script of a more generic target language. This approach saves the acoustic model from retraining to span over a larger space and can easily be extended to multiple languages. A two-stage transliteration process realizes this approach and proves to minimize speech-class confusion. We performed experiments with an end-to-end multilingual speech recognition system for two Indic Languages, namely Nepali and Telugu. The original grapheme space of these languages is projected to the Devanagari script. We achieved a relative reduction of 20% in the Word Error Rate (WER) and 24% in the Character Error Rate (CER) in the transliterated space, over other language-dependent modeling methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.14709v1-abstract-full').style.display = 'none'; document.getElementById('2410.14709v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.12045">arXiv:2410.12045</a> <span> [<a href="https://arxiv.org/pdf/2410.12045">pdf</a>, <a href="https://arxiv.org/format/2410.12045">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="Data Structures and Algorithms">cs.DS</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="Social and Information Networks">cs.SI</span> </div> </div> <p class="title is-5 mathjax"> Differential Privacy on Trust Graphs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Wang%2C+S">Serena Wang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.12045v1-abstract-short" style="display: inline;"> We study differential privacy (DP) in a multi-party setting where each party only trusts a (known) subset of the other parties with its data. Specifically, given a trust graph where vertices correspond to parties and neighbors are mutually trusting, we give a DP algorithm for aggregation with a much better privacy-utility trade-off than in the well-studied local model of DP (where each party trust… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12045v1-abstract-full').style.display = 'inline'; document.getElementById('2410.12045v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.12045v1-abstract-full" style="display: none;"> We study differential privacy (DP) in a multi-party setting where each party only trusts a (known) subset of the other parties with its data. Specifically, given a trust graph where vertices correspond to parties and neighbors are mutually trusting, we give a DP algorithm for aggregation with a much better privacy-utility trade-off than in the well-studied local model of DP (where each party trusts no other party). We further study a robust variant where each party trusts all but an unknown subset of at most $t$ of its neighbors (where $t$ is a given parameter), and give an algorithm for this setting. We complement our algorithms with lower bounds, and discuss implications of our work to other tasks in private learning and analytics. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.12045v1-abstract-full').style.display = 'none'; document.getElementById('2410.12045v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.11097">arXiv:2410.11097</a> <span> [<a href="https://arxiv.org/pdf/2410.11097">pdf</a>, <a href="https://arxiv.org/format/2410.11097">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Sound">cs.SD</span> </div> </div> <p class="title is-5 mathjax"> DMDSpeech: Distilled Diffusion Model Surpassing The Teacher in Zero-shot Speech Synthesis via Direct Metric Optimization </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Li%2C+Y+A">Yingahao Aaron Li</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rithesh Kumar</a>, <a href="/search/cs?searchtype=author&query=Jin%2C+Z">Zeyu Jin</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.11097v1-abstract-short" style="display: inline;"> Diffusion models have demonstrated significant potential in speech synthesis tasks, including text-to-speech (TTS) and voice cloning. However, their iterative denoising processes are inefficient and hinder the application of end-to-end optimization with perceptual metrics. In this paper, we propose a novel method of distilling TTS diffusion models with direct end-to-end evaluation metric optimizat… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.11097v1-abstract-full').style.display = 'inline'; document.getElementById('2410.11097v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.11097v1-abstract-full" style="display: none;"> Diffusion models have demonstrated significant potential in speech synthesis tasks, including text-to-speech (TTS) and voice cloning. However, their iterative denoising processes are inefficient and hinder the application of end-to-end optimization with perceptual metrics. In this paper, we propose a novel method of distilling TTS diffusion models with direct end-to-end evaluation metric optimization, achieving state-of-the-art performance. By incorporating Connectionist Temporal Classification (CTC) loss and Speaker Verification (SV) loss, our approach optimizes perceptual evaluation metrics, leading to notable improvements in word error rate and speaker similarity. Our experiments show that DMDSpeech consistently surpasses prior state-of-the-art models in both naturalness and speaker similarity while being significantly faster. Moreover, our synthetic speech has a higher level of voice similarity to the prompt than the ground truth in both human evaluation and objective speaker similarity metric. This work highlights the potential of direct metric optimization in speech synthesis, allowing models to better align with human auditory preferences. The audio samples are available at https://dmdspeech.github.io/. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.11097v1-abstract-full').style.display = 'none'; document.getElementById('2410.11097v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 14 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.11080">arXiv:2410.11080</a> <span> [<a href="https://arxiv.org/pdf/2410.11080">pdf</a>, <a href="https://arxiv.org/format/2410.11080">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="Graphics">cs.GR</span> </div> </div> <p class="title is-5 mathjax"> Few-shot Novel View Synthesis using Depth Aware 3D Gaussian Splatting </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Raja Kumar</a>, <a href="/search/cs?searchtype=author&query=Vats%2C+V">Vanshika Vats</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.11080v1-abstract-short" style="display: inline;"> 3D Gaussian splatting has surpassed neural radiance field methods in novel view synthesis by achieving lower computational costs and real-time high-quality rendering. Although it produces a high-quality rendering with a lot of input views, its performance drops significantly when only a few views are available. In this work, we address this by proposing a depth-aware Gaussian splatting method for… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.11080v1-abstract-full').style.display = 'inline'; document.getElementById('2410.11080v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.11080v1-abstract-full" style="display: none;"> 3D Gaussian splatting has surpassed neural radiance field methods in novel view synthesis by achieving lower computational costs and real-time high-quality rendering. Although it produces a high-quality rendering with a lot of input views, its performance drops significantly when only a few views are available. In this work, we address this by proposing a depth-aware Gaussian splatting method for few-shot novel view synthesis. We use monocular depth prediction as a prior, along with a scale-invariant depth loss, to constrain the 3D shape under just a few input views. We also model color using lower-order spherical harmonics to avoid overfitting. Further, we observe that removing splats with lower opacity periodically, as performed in the original work, leads to a very sparse point cloud and, hence, a lower-quality rendering. To mitigate this, we retain all the splats, leading to a better reconstruction in a few view settings. Experimental results show that our method outperforms the traditional 3D Gaussian splatting methods by achieving improvements of 10.5% in peak signal-to-noise ratio, 6% in structural similarity index, and 14.1% in perceptual similarity, thereby validating the effectiveness of our approach. The code will be made available at: https://github.com/raja-kumar/depth-aware-3DGS <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.11080v1-abstract-full').style.display = 'none'; document.getElementById('2410.11080v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 14 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Presented in ECCV 2024 workshop S3DSGR</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.10887">arXiv:2410.10887</a> <span> [<a href="https://arxiv.org/pdf/2410.10887">pdf</a>, <a href="https://arxiv.org/format/2410.10887">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="Neural and Evolutionary Computing">cs.NE</span> </div> </div> <p class="title is-5 mathjax"> ActNAS : Generating Efficient YOLO Models using Activation NAS </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sah%2C+S">Sudhakar Sah</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravish Kumar</a>, <a href="/search/cs?searchtype=author&query=Ganji%2C+D+C">Darshan C. Ganji</a>, <a href="/search/cs?searchtype=author&query=Saboori%2C+E">Ehsan Saboori</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.10887v2-abstract-short" style="display: inline;"> Activation functions introduce non-linearity into Neural Networks, enabling them to learn complex patterns. Different activation functions vary in speed and accuracy, ranging from faster but less accurate options like ReLU to slower but more accurate functions like SiLU or SELU. Typically, same activation function is used throughout an entire model architecture. In this paper, we conduct a compreh… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.10887v2-abstract-full').style.display = 'inline'; document.getElementById('2410.10887v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.10887v2-abstract-full" style="display: none;"> Activation functions introduce non-linearity into Neural Networks, enabling them to learn complex patterns. Different activation functions vary in speed and accuracy, ranging from faster but less accurate options like ReLU to slower but more accurate functions like SiLU or SELU. Typically, same activation function is used throughout an entire model architecture. In this paper, we conduct a comprehensive study on the effects of using mixed activation functions in YOLO-based models, evaluating their impact on latency, memory usage, and accuracy across CPU, NPU, and GPU edge devices. We also propose a novel approach that leverages Neural Architecture Search (NAS) to design YOLO models with optimized mixed activation functions.The best model generated through this method demonstrates a slight improvement in mean Average Precision (mAP) compared to baseline model (SiLU), while it is 22.28% faster and consumes 64.15% less memory on the reference NPU device. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.10887v2-abstract-full').style.display = 'none'; document.getElementById('2410.10887v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 14 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 11 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">7 pages, 4 figures, FITML workshop, NeuRIPS 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.09591">arXiv:2410.09591</a> <span> [<a href="https://arxiv.org/pdf/2410.09591">pdf</a>, <a href="https://arxiv.org/format/2410.09591">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"> Unlearn and Burn: Adversarial Machine Unlearning Requests Destroy Model Accuracy </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yangsibo Huang</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+D">Daogao Liu</a>, <a href="/search/cs?searchtype=author&query=Chua%2C+L">Lynn Chua</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Nasr%2C+M">Milad Nasr</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Amer Sinha</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan Zhang</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.09591v1-abstract-short" style="display: inline;"> Machine unlearning algorithms, designed for selective removal of training data from models, have emerged as a promising approach to growing privacy concerns. In this work, we expose a critical yet underexplored vulnerability in the deployment of unlearning systems: the assumption that the data requested for removal is always part of the original training set. We present a threat model where an att… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09591v1-abstract-full').style.display = 'inline'; document.getElementById('2410.09591v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.09591v1-abstract-full" style="display: none;"> Machine unlearning algorithms, designed for selective removal of training data from models, have emerged as a promising approach to growing privacy concerns. In this work, we expose a critical yet underexplored vulnerability in the deployment of unlearning systems: the assumption that the data requested for removal is always part of the original training set. We present a threat model where an attacker can degrade model accuracy by submitting adversarial unlearning requests for data not present in the training set. We propose white-box and black-box attack algorithms and evaluate them through a case study on image classification tasks using the CIFAR-10 and ImageNet datasets, targeting a family of widely used unlearning methods. Our results show extremely poor test accuracy following the attack: 3.6% on CIFAR-10 and 0.4% on ImageNet for white-box attacks, and 8.5% on CIFAR-10 and 1.3% on ImageNet for black-box attacks. Additionally, we evaluate various verification mechanisms to detect the legitimacy of unlearning requests and reveal the challenges in verification, as most of the mechanisms fail to detect stealthy attacks without severely impairing their ability to process valid requests. These findings underscore the urgent need for research on more robust request verification methods and unlearning protocols, should the deployment of machine unlearning systems become more prevalent in the future. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09591v1-abstract-full').style.display = 'none'; document.getElementById('2410.09591v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 12 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.09324">arXiv:2410.09324</a> <span> [<a href="https://arxiv.org/pdf/2410.09324">pdf</a>, <a href="https://arxiv.org/format/2410.09324">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Token Pruning using a Lightweight Background Aware Vision Transformer </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sah%2C+S">Sudhakar Sah</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravish Kumar</a>, <a href="/search/cs?searchtype=author&query=Rohmetra%2C+H">Honnesh Rohmetra</a>, <a href="/search/cs?searchtype=author&query=Saboori%2C+E">Ehsan Saboori</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.09324v1-abstract-short" style="display: inline;"> High runtime memory and high latency puts significant constraint on Vision Transformer training and inference, especially on edge devices. Token pruning reduces the number of input tokens to the ViT based on importance criteria of each token. We present a Background Aware Vision Transformer (BAViT) model, a pre-processing block to object detection models like DETR/YOLOS aimed to reduce runtime mem… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09324v1-abstract-full').style.display = 'inline'; document.getElementById('2410.09324v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.09324v1-abstract-full" style="display: none;"> High runtime memory and high latency puts significant constraint on Vision Transformer training and inference, especially on edge devices. Token pruning reduces the number of input tokens to the ViT based on importance criteria of each token. We present a Background Aware Vision Transformer (BAViT) model, a pre-processing block to object detection models like DETR/YOLOS aimed to reduce runtime memory and increase throughput by using a novel approach to identify background tokens in the image. The background tokens can be pruned completely or partially before feeding to a ViT based object detector. We use the semantic information provided by segmentation map and/or bounding box annotation to train a few layers of ViT to classify tokens to either foreground or background. Using 2 layers and 10 layers of BAViT, background and foreground tokens can be separated with 75% and 88% accuracy on VOC dataset and 71% and 80% accuracy on COCO dataset respectively. We show a 2 layer BAViT-small model as pre-processor to YOLOS can increase the throughput by 30% - 40% with a mAP drop of 3% without any sparse fine-tuning and 2% with sparse fine-tuning. Our approach is specifically targeted for Edge AI use cases. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.09324v1-abstract-full').style.display = 'none'; document.getElementById('2410.09324v1-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">7 pages, 2 tables, 4 figures, FITML workshop@NeuRIPS 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.06954">arXiv:2410.06954</a> <span> [<a href="https://arxiv.org/pdf/2410.06954">pdf</a>, <a href="https://arxiv.org/format/2410.06954">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> </div> <div 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/10.56553/popets-2025-0038">10.56553/popets-2025-0038 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> How Unique is Whose Web Browser? The role of demographics in browser fingerprinting among US users </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Berke%2C+A">Alex Berke</a>, <a href="/search/cs?searchtype=author&query=Bacis%2C+E">Enrico Bacis</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Lassonde%2C+R">Robin Lassonde</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Syed%2C+U">Umar Syed</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.06954v3-abstract-short" style="display: inline;"> Browser fingerprinting can be used to identify and track users across the Web, even without cookies, by collecting attributes from users' devices to create unique "fingerprints". This technique and resulting privacy risks have been studied for over a decade. Yet further research is limited because prior studies used data not publicly available. Additionally, data in prior studies lacked user demog… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06954v3-abstract-full').style.display = 'inline'; document.getElementById('2410.06954v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.06954v3-abstract-full" style="display: none;"> Browser fingerprinting can be used to identify and track users across the Web, even without cookies, by collecting attributes from users' devices to create unique "fingerprints". This technique and resulting privacy risks have been studied for over a decade. Yet further research is limited because prior studies used data not publicly available. Additionally, data in prior studies lacked user demographics. Here we provide a first-of-its-kind dataset to enable further research. It includes browser attributes with users' demographics and survey responses, collected with informed consent from 8,400 US study participants. We use this dataset to demonstrate how fingerprinting risks differ across demographic groups. For example, we find lower income users are more at risk, and find that as users' age increases, they are both more likely to be concerned about fingerprinting and at real risk of fingerprinting. Furthermore, we demonstrate an overlooked risk: user demographics, such as gender, age, income level and race, can be inferred from browser attributes commonly used for fingerprinting, and we identify which browser attributes most contribute to this risk. Our data collection process also conducted an experiment to study what impacts users' likelihood to share browser data for open research, in order to inform future data collection efforts, with responses from 12,461 total participants. Female participants were significantly less likely to share their browser data, as were participants who were shown the browser data we asked to collect. Overall, we show the important role of user demographics in the ongoing work that intends to assess fingerprinting risks and improve user privacy, with findings to inform future privacy enhancing browser developments. The dataset and data collection tool we provide can be used to further study research questions not addressed in this work. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06954v3-abstract-full').style.display = 'none'; document.getElementById('2410.06954v3-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 18 November, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 9 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">In Proceedings on Privacy Enhancing Technologies 2025(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/2410.06008">arXiv:2410.06008</a> <span> [<a href="https://arxiv.org/pdf/2410.06008">pdf</a>, <a href="https://arxiv.org/format/2410.06008">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"> Sitting, Standing and Walking Control of the Series-Parallel Hybrid Recupera-Reha Exoskeleton </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Tijjani%2C+I">Ibrahim Tijjani</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rohit Kumar</a>, <a href="/search/cs?searchtype=author&query=Boukheddimi%2C+M">Melya Boukheddimi</a>, <a href="/search/cs?searchtype=author&query=Trampler%2C+M">Mathias Trampler</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S">Shivesh Kumar</a>, <a href="/search/cs?searchtype=author&query=Kirchner%2C+F">Frank Kirchner</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.06008v1-abstract-short" style="display: inline;"> This paper presents advancements in the functionalities of the Recupera-Reha lower extremity exoskeleton robot. The exoskeleton features a series-parallel hybrid design characterized by multiple kinematic loops resulting in 148 degrees of freedom in its spanning tree and 102 independent loop closure constraints, which poses significant challenges for modeling and control. To address these challeng… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06008v1-abstract-full').style.display = 'inline'; document.getElementById('2410.06008v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.06008v1-abstract-full" style="display: none;"> This paper presents advancements in the functionalities of the Recupera-Reha lower extremity exoskeleton robot. The exoskeleton features a series-parallel hybrid design characterized by multiple kinematic loops resulting in 148 degrees of freedom in its spanning tree and 102 independent loop closure constraints, which poses significant challenges for modeling and control. To address these challenges, we applied an optimal control approach to generate feasible trajectories such as sitting, standing, and static walking, and tested these trajectories on the exoskeleton robot. Our method efficiently solves the optimal control problem using a serial abstraction of the model to generate trajectories. It then utilizes the full series-parallel hybrid model, which takes all the kinematic loop constraints into account to generate the final actuator commands. The experimental results demonstrate the effectiveness of our approach in generating the desired motions for the exoskeleton. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.06008v1-abstract-full').style.display = 'none'; document.getElementById('2410.06008v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 8 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> <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">8 pages, 16 figures, IEEE-RAS International Conference on Humanoid Robots 2024</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">MSC Class:</span> 68-06 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.02828">arXiv:2410.02828</a> <span> [<a href="https://arxiv.org/pdf/2410.02828">pdf</a>, <a href="https://arxiv.org/format/2410.02828">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> </div> </div> <p class="title is-5 mathjax"> PyRIT: A Framework for Security Risk Identification and Red Teaming in Generative AI System </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Munoz%2C+G+D+L">Gary D. Lopez Munoz</a>, <a href="/search/cs?searchtype=author&query=Minnich%2C+A+J">Amanda J. Minnich</a>, <a href="/search/cs?searchtype=author&query=Lutz%2C+R">Roman Lutz</a>, <a href="/search/cs?searchtype=author&query=Lundeen%2C+R">Richard Lundeen</a>, <a href="/search/cs?searchtype=author&query=Dheekonda%2C+R+S+R">Raja Sekhar Rao Dheekonda</a>, <a href="/search/cs?searchtype=author&query=Chikanov%2C+N">Nina Chikanov</a>, <a href="/search/cs?searchtype=author&query=Jagdagdorj%2C+B">Bolor-Erdene Jagdagdorj</a>, <a href="/search/cs?searchtype=author&query=Pouliot%2C+M">Martin Pouliot</a>, <a href="/search/cs?searchtype=author&query=Chawla%2C+S">Shiven Chawla</a>, <a href="/search/cs?searchtype=author&query=Maxwell%2C+W">Whitney Maxwell</a>, <a href="/search/cs?searchtype=author&query=Bullwinkel%2C+B">Blake Bullwinkel</a>, <a href="/search/cs?searchtype=author&query=Pratt%2C+K">Katherine Pratt</a>, <a href="/search/cs?searchtype=author&query=de+Gruyter%2C+J">Joris de Gruyter</a>, <a href="/search/cs?searchtype=author&query=Siska%2C+C">Charlotte Siska</a>, <a href="/search/cs?searchtype=author&query=Bryan%2C+P">Pete Bryan</a>, <a href="/search/cs?searchtype=author&query=Westerhoff%2C+T">Tori Westerhoff</a>, <a href="/search/cs?searchtype=author&query=Kawaguchi%2C+C">Chang Kawaguchi</a>, <a href="/search/cs?searchtype=author&query=Seifert%2C+C">Christian Seifert</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R+S+S">Ram Shankar Siva Kumar</a>, <a href="/search/cs?searchtype=author&query=Zunger%2C+Y">Yonatan Zunger</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.02828v1-abstract-short" style="display: inline;"> Generative Artificial Intelligence (GenAI) is becoming ubiquitous in our daily lives. The increase in computational power and data availability has led to a proliferation of both single- and multi-modal models. As the GenAI ecosystem matures, the need for extensible and model-agnostic risk identification frameworks is growing. To meet this need, we introduce the Python Risk Identification Toolkit… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.02828v1-abstract-full').style.display = 'inline'; document.getElementById('2410.02828v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.02828v1-abstract-full" style="display: none;"> Generative Artificial Intelligence (GenAI) is becoming ubiquitous in our daily lives. The increase in computational power and data availability has led to a proliferation of both single- and multi-modal models. As the GenAI ecosystem matures, the need for extensible and model-agnostic risk identification frameworks is growing. To meet this need, we introduce the Python Risk Identification Toolkit (PyRIT), an open-source framework designed to enhance red teaming efforts in GenAI systems. PyRIT is a model- and platform-agnostic tool that enables red teamers to probe for and identify novel harms, risks, and jailbreaks in multimodal generative AI models. Its composable architecture facilitates the reuse of core building blocks and allows for extensibility to future models and modalities. This paper details the challenges specific to red teaming generative AI systems, the development and features of PyRIT, and its practical applications in real-world scenarios. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.02828v1-abstract-full').style.display = 'none'; document.getElementById('2410.02828v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 1 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2410.00019">arXiv:2410.00019</a> <span> [<a href="https://arxiv.org/pdf/2410.00019">pdf</a>, <a href="https://arxiv.org/ps/2410.00019">ps</a>, <a href="https://arxiv.org/format/2410.00019">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computational Engineering, Finance, and Science">cs.CE</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Dynamical Systems">math.DS</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Numerical Analysis">math.NA</span> </div> </div> <p class="title is-5 mathjax"> Hybridized Projected Differential Transform Method For collisional-breakage equation </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Shweta"> Shweta</a>, <a href="/search/cs?searchtype=author&query=Hussain%2C+S">Saddam Hussain</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rajesh Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2410.00019v1-abstract-short" style="display: inline;"> The non-linear collision induced fragmentation plays a crucial role in modeling several engineering and physical problems. In contrast to linear breakage, it has not been thoroughly investigated in the existing literature. This study introduces an innovative method that leverages the Elzaki integral transform as a preparatory step to enhance the accuracy and convergence of domain decomposition, us… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.00019v1-abstract-full').style.display = 'inline'; document.getElementById('2410.00019v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2410.00019v1-abstract-full" style="display: none;"> The non-linear collision induced fragmentation plays a crucial role in modeling several engineering and physical problems. In contrast to linear breakage, it has not been thoroughly investigated in the existing literature. This study introduces an innovative method that leverages the Elzaki integral transform as a preparatory step to enhance the accuracy and convergence of domain decomposition, used alongside the projected differential transform method to obtain closed-form or series approximations of solutions for the collisional breakage equation (CBE). A significant advantages of this technique is its capability to directly address both linear and nonlinear differential equations without the need for discretization or linearization. The mathematical framework is reinforced by a thorough convergence analysis, applying fixed point theory within an adequately defined Banach space. Additionally, error estimates for the approximated solutions are derived, offering more profound insights into the accuracy and dependability of the proposed method. The validity of this approach is demonstrated by comparing the obtained results with exact or finite volume approximated solutions considering several physical examples. Interestingly, the proposed algorithm yields accurate approximations for the number density functions as well as moments with fewer terms and maintains higher precision over extended time periods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2410.00019v1-abstract-full').style.display = 'none'; document.getElementById('2410.00019v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> October 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.18866">arXiv:2409.18866</a> <span> [<a href="https://arxiv.org/pdf/2409.18866">pdf</a>, <a href="https://arxiv.org/format/2409.18866">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"> MCUBench: A Benchmark of Tiny Object Detectors on MCUs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sah%2C+S">Sudhakar Sah</a>, <a href="/search/cs?searchtype=author&query=Ganji%2C+D+C">Darshan C. Ganji</a>, <a href="/search/cs?searchtype=author&query=Grimaldi%2C+M">Matteo Grimaldi</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravish Kumar</a>, <a href="/search/cs?searchtype=author&query=Hoffman%2C+A">Alexander Hoffman</a>, <a href="/search/cs?searchtype=author&query=Rohmetra%2C+H">Honnesh Rohmetra</a>, <a href="/search/cs?searchtype=author&query=Saboori%2C+E">Ehsan Saboori</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2409.18866v1-abstract-short" style="display: inline;"> We introduce MCUBench, a benchmark featuring over 100 YOLO-based object detection models evaluated on the VOC dataset across seven different MCUs. This benchmark provides detailed data on average precision, latency, RAM, and Flash usage for various input resolutions and YOLO-based one-stage detectors. By conducting a controlled comparison with a fixed training pipeline, we collect comprehensive pe… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.18866v1-abstract-full').style.display = 'inline'; document.getElementById('2409.18866v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.18866v1-abstract-full" style="display: none;"> We introduce MCUBench, a benchmark featuring over 100 YOLO-based object detection models evaluated on the VOC dataset across seven different MCUs. This benchmark provides detailed data on average precision, latency, RAM, and Flash usage for various input resolutions and YOLO-based one-stage detectors. By conducting a controlled comparison with a fixed training pipeline, we collect comprehensive performance metrics. Our Pareto-optimal analysis shows that integrating modern detection heads and training techniques allows various YOLO architectures, including legacy models like YOLOv3, to achieve a highly efficient tradeoff between mean Average Precision (mAP) and latency. MCUBench serves as a valuable tool for benchmarking the MCU performance of contemporary object detectors and aids in model selection based on specific constraints. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.18866v1-abstract-full').style.display = 'none'; document.getElementById('2409.18866v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 27 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> <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">Code and data are available at https://github.com/Deeplite/deeplite-torch-zoo</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.18193">arXiv:2409.18193</a> <span> [<a href="https://arxiv.org/pdf/2409.18193">pdf</a>, <a href="https://arxiv.org/format/2409.18193">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"> LowREm: A Repository of Word Embeddings for 87 Low-Resource Languages Enhanced with Multilingual Graph Knowledge </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Gurgurov%2C+D">Daniil Gurgurov</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rishu Kumar</a>, <a href="/search/cs?searchtype=author&query=Ostermann%2C+S">Simon Ostermann</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2409.18193v1-abstract-short" style="display: inline;"> Contextualized embeddings based on large language models (LLMs) are available for various languages, but their coverage is often limited for lower resourced languages. Training LLMs for such languages is often difficult due to insufficient data and high computational cost. Especially for very low resource languages, static word embeddings thus still offer a viable alternative. There is, however, a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.18193v1-abstract-full').style.display = 'inline'; document.getElementById('2409.18193v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.18193v1-abstract-full" style="display: none;"> Contextualized embeddings based on large language models (LLMs) are available for various languages, but their coverage is often limited for lower resourced languages. Training LLMs for such languages is often difficult due to insufficient data and high computational cost. Especially for very low resource languages, static word embeddings thus still offer a viable alternative. There is, however, a notable lack of comprehensive repositories with such embeddings for diverse languages. To address this, we present LowREm, a centralized repository of static embeddings for 87 low-resource languages. We also propose a novel method to enhance GloVe-based embeddings by integrating multilingual graph knowledge, utilizing another source of knowledge. We demonstrate the superior performance of our enhanced embeddings as compared to contextualized embeddings extracted from XLM-R on sentiment analysis. Our code and data are publicly available under https://huggingface.co/DFKI. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.18193v1-abstract-full').style.display = 'none'; document.getElementById('2409.18193v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 26 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Short paper, preview</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.17777">arXiv:2409.17777</a> <span> [<a href="https://arxiv.org/pdf/2409.17777">pdf</a>, <a href="https://arxiv.org/format/2409.17777">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computer Vision and Pattern Recognition">cs.CV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Harnessing Shared Relations via Multimodal Mixup Contrastive Learning for Multimodal Classification </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Raja Kumar</a>, <a href="/search/cs?searchtype=author&query=Singhal%2C+R">Raghav Singhal</a>, <a href="/search/cs?searchtype=author&query=Kulkarni%2C+P">Pranamya Kulkarni</a>, <a href="/search/cs?searchtype=author&query=Mehta%2C+D">Deval Mehta</a>, <a href="/search/cs?searchtype=author&query=Jadhav%2C+K">Kshitij Jadhav</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2409.17777v2-abstract-short" style="display: inline;"> Deep multimodal learning has shown remarkable success by leveraging contrastive learning to capture explicit one-to-one relations across modalities. However, real-world data often exhibits shared relations beyond simple pairwise associations. We propose M3CoL, a Multimodal Mixup Contrastive Learning approach to capture nuanced shared relations inherent in multimodal data. Our key contribution is a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17777v2-abstract-full').style.display = 'inline'; document.getElementById('2409.17777v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.17777v2-abstract-full" style="display: none;"> Deep multimodal learning has shown remarkable success by leveraging contrastive learning to capture explicit one-to-one relations across modalities. However, real-world data often exhibits shared relations beyond simple pairwise associations. We propose M3CoL, a Multimodal Mixup Contrastive Learning approach to capture nuanced shared relations inherent in multimodal data. Our key contribution is a Mixup-based contrastive loss that learns robust representations by aligning mixed samples from one modality with their corresponding samples from other modalities thereby capturing shared relations between them. For multimodal classification tasks, we introduce a framework that integrates a fusion module with unimodal prediction modules for auxiliary supervision during training, complemented by our proposed Mixup-based contrastive loss. Through extensive experiments on diverse datasets (N24News, ROSMAP, BRCA, and Food-101), we demonstrate that M3CoL effectively captures shared multimodal relations and generalizes across domains. It outperforms state-of-the-art methods on N24News, ROSMAP, and BRCA, while achieving comparable performance on Food-101. Our work highlights the significance of learning shared relations for robust multimodal learning, opening up promising avenues for future research. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.17777v2-abstract-full').style.display = 'none'; document.getElementById('2409.17777v2-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 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 26 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">RK and RS contributed equally to this work, 20 Pages, 8 Figures, 9 Tables. Another version of the paper accepted at NeurIPS 2024 Workshop on Unifying Representations in Neural Models (UniReps)</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2409.08507">arXiv:2409.08507</a> <span> [<a href="https://arxiv.org/pdf/2409.08507">pdf</a>, <a href="https://arxiv.org/format/2409.08507">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Systems and Control">eess.SY</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Dynamical Systems">math.DS</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Optimization and Control">math.OC</span> </div> </div> <p class="title is-5 mathjax"> Three-dimensional Nonlinear Path-following Guidance with Bounded Input Constraints </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+S">Saurabh Kumar</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S+R">Shashi Ranjan Kumar</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Abhinav Sinha</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2409.08507v1-abstract-short" style="display: inline;"> In this paper, we consider the tracking of arbitrary curvilinear geometric paths in three-dimensional output spaces of unmanned aerial vehicles (UAVs) without pre-specified timing requirements, commonly referred to as path-following problems, subjected to bounded inputs. Specifically, we propose a novel nonlinear path-following guidance law for a UAV that enables it to follow any smooth curvilinea… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.08507v1-abstract-full').style.display = 'inline'; document.getElementById('2409.08507v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2409.08507v1-abstract-full" style="display: none;"> In this paper, we consider the tracking of arbitrary curvilinear geometric paths in three-dimensional output spaces of unmanned aerial vehicles (UAVs) without pre-specified timing requirements, commonly referred to as path-following problems, subjected to bounded inputs. Specifically, we propose a novel nonlinear path-following guidance law for a UAV that enables it to follow any smooth curvilinear path in three dimensions while accounting for the bounded control authority in the design. The proposed solution offers a general treatment of the path-following problem by removing the dependency on the path's geometry, which makes it applicable to paths with varying levels of complexity and smooth curvatures. Additionally, the proposed strategy draws inspiration from the pursuit guidance approach, which is known for its simplicity and ease of implementation. Theoretical analysis guarantees that the UAV converges to its desired path within a fixed time and remains on it irrespective of its initial configuration with respect to the path. Finally, the simulations demonstrate the merits and effectiveness of the proposed guidance strategy through a wide range of engagement scenarios, showcasing the UAV's ability to follow diverse curvilinear paths accurately. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2409.08507v1-abstract-full').style.display = 'none'; document.getElementById('2409.08507v1-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> September 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.16599">arXiv:2408.16599</a> <span> [<a href="https://arxiv.org/pdf/2408.16599">pdf</a>, <a href="https://arxiv.org/format/2408.16599">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"> sEMG-Driven Physics-Informed Gated Recurrent Networks for Modeling Upper Limb Multi-Joint Movement Dynamics </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rajnish Kumar</a>, <a href="/search/cs?searchtype=author&query=Gupta%2C+A">Anand Gupta</a>, <a href="/search/cs?searchtype=author&query=Muthukrishnan%2C+S+P">Suriya Prakash Muthukrishnan</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+L">Lalan Kumar</a>, <a href="/search/cs?searchtype=author&query=Roy%2C+S">Sitikantha Roy</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2408.16599v1-abstract-short" style="display: inline;"> Exoskeletons and rehabilitation systems offer great potential for enhancing human strength and recovery through advanced human-machine interfaces (HMIs) that adapt to movement dynamics. However, the real-time application of physics-informed neural networks (PINNs) is limited by their reliance on fixed input lengths and surrogate models. This study introduces a novel physics-informed Gated Recurren… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.16599v1-abstract-full').style.display = 'inline'; document.getElementById('2408.16599v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.16599v1-abstract-full" style="display: none;"> Exoskeletons and rehabilitation systems offer great potential for enhancing human strength and recovery through advanced human-machine interfaces (HMIs) that adapt to movement dynamics. However, the real-time application of physics-informed neural networks (PINNs) is limited by their reliance on fixed input lengths and surrogate models. This study introduces a novel physics-informed Gated Recurrent Network (PiGRN) designed to predict multi-joint torques using surface electromyography (sEMG) data. The PiGRN model employs a Gated Recurrent Unit (GRU) to convert time-series sEMG inputs into multi-joint kinematics and external loads, which are then integrated into an equation of motion to ensure consistency with physical laws. Experimental validation with sEMG data from five participants performing elbow flexion-extension tasks showed that the PiGRN model accurately predicted joint torques for 10 unfamiliar movements, with RMSE values between 4.02\% and 11.40\% and correlation coefficients ranging from 0.87 to 0.98. These findings highlight the PiGRN's potential for real-time exoskeleton and rehabilitation applications. Future research will explore more diverse datasets, improve musculoskeletal models, and investigate unsupervised learning methods. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.16599v1-abstract-full').style.display = 'none'; document.getElementById('2408.16599v1-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 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.06113">arXiv:2408.06113</a> <span> [<a href="https://arxiv.org/pdf/2408.06113">pdf</a>, <a href="https://arxiv.org/format/2408.06113">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"> IIT Bombay Racing Driverless: Autonomous Driving Stack for Formula Student AI </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Rampuria%2C+Y">Yash Rampuria</a>, <a href="/search/cs?searchtype=author&query=Boliya%2C+D">Deep Boliya</a>, <a href="/search/cs?searchtype=author&query=Gupta%2C+S">Shreyash Gupta</a>, <a href="/search/cs?searchtype=author&query=Iyengar%2C+G">Gopalan Iyengar</a>, <a href="/search/cs?searchtype=author&query=Rohilla%2C+A">Ayush Rohilla</a>, <a href="/search/cs?searchtype=author&query=Vyas%2C+M">Mohak Vyas</a>, <a href="/search/cs?searchtype=author&query=Langde%2C+C">Chaitanya Langde</a>, <a href="/search/cs?searchtype=author&query=Chanda%2C+M+V">Mehul Vijay Chanda</a>, <a href="/search/cs?searchtype=author&query=Matai%2C+R+G">Ronak Gautam Matai</a>, <a href="/search/cs?searchtype=author&query=Namitha%2C+K">Kothapalli Namitha</a>, <a href="/search/cs?searchtype=author&query=Pawar%2C+A">Ajinkya Pawar</a>, <a href="/search/cs?searchtype=author&query=Biswas%2C+B">Bhaskar Biswas</a>, <a href="/search/cs?searchtype=author&query=Agarwal%2C+N">Nakul Agarwal</a>, <a href="/search/cs?searchtype=author&query=Khandelwal%2C+R">Rajit Khandelwal</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rohan Kumar</a>, <a href="/search/cs?searchtype=author&query=Agarwal%2C+S">Shubham Agarwal</a>, <a href="/search/cs?searchtype=author&query=Patel%2C+V">Vishwam Patel</a>, <a href="/search/cs?searchtype=author&query=Rathore%2C+A+S">Abhimanyu Singh Rathore</a>, <a href="/search/cs?searchtype=author&query=Rahman%2C+A">Amna Rahman</a>, <a href="/search/cs?searchtype=author&query=Mishra%2C+A">Ayush Mishra</a>, <a href="/search/cs?searchtype=author&query=Tangri%2C+Y">Yash Tangri</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2408.06113v1-abstract-short" style="display: inline;"> This work presents the design and development of IIT Bombay Racing's Formula Student style autonomous racecar algorithm capable of running at the racing events of Formula Student-AI, held in the UK. The car employs a cutting-edge sensor suite of the compute unit NVIDIA Jetson Orin AGX, 2 ZED2i stereo cameras, 1 Velodyne Puck VLP16 LiDAR and SBG Systems Ellipse N GNSS/INS IMU. It features deep lear… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.06113v1-abstract-full').style.display = 'inline'; document.getElementById('2408.06113v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.06113v1-abstract-full" style="display: none;"> This work presents the design and development of IIT Bombay Racing's Formula Student style autonomous racecar algorithm capable of running at the racing events of Formula Student-AI, held in the UK. The car employs a cutting-edge sensor suite of the compute unit NVIDIA Jetson Orin AGX, 2 ZED2i stereo cameras, 1 Velodyne Puck VLP16 LiDAR and SBG Systems Ellipse N GNSS/INS IMU. It features deep learning algorithms and control systems to navigate complex tracks and execute maneuvers without any human intervention. The design process involved extensive simulations and testing to optimize the vehicle's performance and ensure its safety. The algorithms have been tested on a small scale, in-house manufactured 4-wheeled robot and on simulation software. The results obtained for testing various algorithms in perception, simultaneous localization and mapping, path planning and controls have been detailed. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.06113v1-abstract-full').style.display = 'none'; document.getElementById('2408.06113v1-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 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">8 pages, 19 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/2408.05755">arXiv:2408.05755</a> <span> [<a href="https://arxiv.org/pdf/2408.05755">pdf</a>, <a href="https://arxiv.org/format/2408.05755">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Social and Information Networks">cs.SI</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/10.1007/s13278-024-01331-9">10.1007/s13278-024-01331-9 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> Effect of Perturbation and Topological Structure on Synchronization Dynamics in Multilayer Networks </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rajesh Kumar</a>, <a href="/search/cs?searchtype=author&query=Kumari%2C+S">Suchi Kumari</a>, <a href="/search/cs?searchtype=author&query=Mishra%2C+A">Anubhav Mishra</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2408.05755v1-abstract-short" style="display: inline;"> The way the topological structure transforms from a decoupled to a coupled state in multiplex networks has been extensively studied through both analytical and numerical approaches, often utilizing models of artificial networks. These studies typically assume uniform interconnections between layers to simplify the analytical treatment of structural properties in multiplex networks. However, this a… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.05755v1-abstract-full').style.display = 'inline'; document.getElementById('2408.05755v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.05755v1-abstract-full" style="display: none;"> The way the topological structure transforms from a decoupled to a coupled state in multiplex networks has been extensively studied through both analytical and numerical approaches, often utilizing models of artificial networks. These studies typically assume uniform interconnections between layers to simplify the analytical treatment of structural properties in multiplex networks. However, this assumption is not applicable for real networks, where the heterogeneity of link weights is an intrinsic characteristic. Therefore, in this paper, link weights are calculated considering the node's reputation and the impact of the inter-layer link weights are assessed on the overall network's structural characteristics. These characteristics include synchronization time, stability of synchronization, and the second-smallest eigenvalue of the Laplacian matrix (algebraic connectivity). Our findings reveal that the perturbation in link weights (intra-layer) causes a transition in the algebraic connectivity whereas variation in inter-layer link weights has a significant impact on the synchronization stability and synchronization time in the multiplex networks. This analysis is different from the predictions made under the assumption of equal inter-layer link weights. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.05755v1-abstract-full').style.display = 'none'; document.getElementById('2408.05755v1-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 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">22 pages, 14 figures, 3 tables</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Report number:</span> vol. 14, no. 179 </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2408.00118">arXiv:2408.00118</a> <span> [<a href="https://arxiv.org/pdf/2408.00118">pdf</a>, <a href="https://arxiv.org/format/2408.00118">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Gemma 2: Improving Open Language Models at a Practical Size </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Gemma+Team"> Gemma Team</a>, <a href="/search/cs?searchtype=author&query=Riviere%2C+M">Morgane Riviere</a>, <a href="/search/cs?searchtype=author&query=Pathak%2C+S">Shreya Pathak</a>, <a href="/search/cs?searchtype=author&query=Sessa%2C+P+G">Pier Giuseppe Sessa</a>, <a href="/search/cs?searchtype=author&query=Hardin%2C+C">Cassidy Hardin</a>, <a href="/search/cs?searchtype=author&query=Bhupatiraju%2C+S">Surya Bhupatiraju</a>, <a href="/search/cs?searchtype=author&query=Hussenot%2C+L">L茅onard Hussenot</a>, <a href="/search/cs?searchtype=author&query=Mesnard%2C+T">Thomas Mesnard</a>, <a href="/search/cs?searchtype=author&query=Shahriari%2C+B">Bobak Shahriari</a>, <a href="/search/cs?searchtype=author&query=Ram%C3%A9%2C+A">Alexandre Ram茅</a>, <a href="/search/cs?searchtype=author&query=Ferret%2C+J">Johan Ferret</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+P">Peter Liu</a>, <a href="/search/cs?searchtype=author&query=Tafti%2C+P">Pouya Tafti</a>, <a href="/search/cs?searchtype=author&query=Friesen%2C+A">Abe Friesen</a>, <a href="/search/cs?searchtype=author&query=Casbon%2C+M">Michelle Casbon</a>, <a href="/search/cs?searchtype=author&query=Ramos%2C+S">Sabela Ramos</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravin Kumar</a>, <a href="/search/cs?searchtype=author&query=Lan%2C+C+L">Charline Le Lan</a>, <a href="/search/cs?searchtype=author&query=Jerome%2C+S">Sammy Jerome</a>, <a href="/search/cs?searchtype=author&query=Tsitsulin%2C+A">Anton Tsitsulin</a>, <a href="/search/cs?searchtype=author&query=Vieillard%2C+N">Nino Vieillard</a>, <a href="/search/cs?searchtype=author&query=Stanczyk%2C+P">Piotr Stanczyk</a>, <a href="/search/cs?searchtype=author&query=Girgin%2C+S">Sertan Girgin</a>, <a href="/search/cs?searchtype=author&query=Momchev%2C+N">Nikola Momchev</a>, <a href="/search/cs?searchtype=author&query=Hoffman%2C+M">Matt Hoffman</a> , et al. (173 additional authors not shown) </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2408.00118v3-abstract-short" style="display: inline;"> In this work, we introduce Gemma 2, a new addition to the Gemma family of lightweight, state-of-the-art open models, ranging in scale from 2 billion to 27 billion parameters. In this new version, we apply several known technical modifications to the Transformer architecture, such as interleaving local-global attentions (Beltagy et al., 2020a) and group-query attention (Ainslie et al., 2023). We al… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.00118v3-abstract-full').style.display = 'inline'; document.getElementById('2408.00118v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2408.00118v3-abstract-full" style="display: none;"> In this work, we introduce Gemma 2, a new addition to the Gemma family of lightweight, state-of-the-art open models, ranging in scale from 2 billion to 27 billion parameters. In this new version, we apply several known technical modifications to the Transformer architecture, such as interleaving local-global attentions (Beltagy et al., 2020a) and group-query attention (Ainslie et al., 2023). We also train the 2B and 9B models with knowledge distillation (Hinton et al., 2015) instead of next token prediction. The resulting models deliver the best performance for their size, and even offer competitive alternatives to models that are 2-3 times bigger. We release all our models to the community. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2408.00118v3-abstract-full').style.display = 'none'; document.getElementById('2408.00118v3-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 2 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 31 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> August 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.21141">arXiv:2407.21141</a> <span> [<a href="https://arxiv.org/pdf/2407.21141">pdf</a>, <a href="https://arxiv.org/format/2407.21141">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"> FL-DECO-BC: A Privacy-Preserving, Provably Secure, and Provenance-Preserving Federated Learning Framework with Decentralized Oracles on Blockchain for VANETs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Narkedimilli%2C+S">Sathwik Narkedimilli</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R+A">Rayachoti Arun Kumar</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+N+V+S">N. V. Saran Kumar</a>, <a href="/search/cs?searchtype=author&query=Reddy%2C+R+P">Ramapathruni Praneeth Reddy</a>, <a href="/search/cs?searchtype=author&query=C%2C+P+K">Pavan Kumar C</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.21141v1-abstract-short" style="display: inline;"> Vehicular Ad-Hoc Networks (VANETs) hold immense potential for improving traffic safety and efficiency. However, traditional centralized approaches for machine learning in VANETs raise concerns about data privacy and security. Federated Learning (FL) offers a solution that enables collaborative model training without sharing raw data. This paper proposes FL-DECO-BC as a novel privacy-preserving, pr… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.21141v1-abstract-full').style.display = 'inline'; document.getElementById('2407.21141v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.21141v1-abstract-full" style="display: none;"> Vehicular Ad-Hoc Networks (VANETs) hold immense potential for improving traffic safety and efficiency. However, traditional centralized approaches for machine learning in VANETs raise concerns about data privacy and security. Federated Learning (FL) offers a solution that enables collaborative model training without sharing raw data. This paper proposes FL-DECO-BC as a novel privacy-preserving, provably secure, and provenance-preserving federated learning framework specifically designed for VANETs. FL-DECO-BC leverages decentralized oracles on blockchain to securely access external data sources while ensuring data privacy through advanced techniques. The framework guarantees provable security through cryptographic primitives and formal verification methods. Furthermore, FL-DECO-BC incorporates a provenance-preserving design to track data origin and history, fostering trust and accountability. This combination of features empowers VANETs with secure and privacy-conscious machine-learning capabilities, paving the way for advanced traffic management and safety applications. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.21141v1-abstract-full').style.display = 'none'; document.getElementById('2407.21141v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 30 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.19048">arXiv:2407.19048</a> <span> [<a href="https://arxiv.org/pdf/2407.19048">pdf</a>, <a href="https://arxiv.org/format/2407.19048">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="General Relativity and Quantum Cosmology">gr-qc</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Instrumentation and Methods for Astrophysics">astro-ph.IM</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"> Rapid Likelihood Free Inference of Compact Binary Coalescences using Accelerated Hardware </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chatterjee%2C+D">Deep Chatterjee</a>, <a href="/search/cs?searchtype=author&query=Marx%2C+E">Ethan Marx</a>, <a href="/search/cs?searchtype=author&query=Benoit%2C+W">William Benoit</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Desai%2C+M">Malina Desai</a>, <a href="/search/cs?searchtype=author&query=Govorkova%2C+E">Ekaterina Govorkova</a>, <a href="/search/cs?searchtype=author&query=Gunny%2C+A">Alec Gunny</a>, <a href="/search/cs?searchtype=author&query=Moreno%2C+E">Eric Moreno</a>, <a href="/search/cs?searchtype=author&query=Omer%2C+R">Rafia Omer</a>, <a href="/search/cs?searchtype=author&query=Raikman%2C+R">Ryan Raikman</a>, <a href="/search/cs?searchtype=author&query=Saleem%2C+M">Muhammed Saleem</a>, <a href="/search/cs?searchtype=author&query=Aggarwal%2C+S">Shrey Aggarwal</a>, <a href="/search/cs?searchtype=author&query=Coughlin%2C+M+W">Michael W. Coughlin</a>, <a href="/search/cs?searchtype=author&query=Harris%2C+P">Philip Harris</a>, <a href="/search/cs?searchtype=author&query=Katsavounidis%2C+E">Erik Katsavounidis</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.19048v1-abstract-short" style="display: inline;"> We report a gravitational-wave parameter estimation algorithm, AMPLFI, based on likelihood-free inference using normalizing flows. The focus of AMPLFI is to perform real-time parameter estimation for candidates detected by machine-learning based compact binary coalescence search, Aframe. We present details of our algorithm and optimizations done related to data-loading and pre-processing on accele… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.19048v1-abstract-full').style.display = 'inline'; document.getElementById('2407.19048v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.19048v1-abstract-full" style="display: none;"> We report a gravitational-wave parameter estimation algorithm, AMPLFI, based on likelihood-free inference using normalizing flows. The focus of AMPLFI is to perform real-time parameter estimation for candidates detected by machine-learning based compact binary coalescence search, Aframe. We present details of our algorithm and optimizations done related to data-loading and pre-processing on accelerated hardware. We train our model using binary black-hole (BBH) simulations on real LIGO-Virgo detector noise. Our model has $\sim 6$ million trainable parameters with training times $\lesssim 24$ hours. Based on online deployment on a mock data stream of LIGO-Virgo data, Aframe + AMPLFI is able to pick up BBH candidates and infer parameters for real-time alerts from data acquisition with a net latency of $\sim 6$s. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.19048v1-abstract-full').style.display = 'none'; document.getElementById('2407.19048v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 26 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Submitted to MLST</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.17712">arXiv:2407.17712</a> <span> [<a href="https://arxiv.org/pdf/2407.17712">pdf</a>, <a href="https://arxiv.org/format/2407.17712">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Data Structures and Algorithms">cs.DS</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"> Improving Online Algorithms via ML Predictions </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Purohit%2C+M">Manish Purohit</a>, <a href="/search/cs?searchtype=author&query=Svitkina%2C+Z">Zoya Svitkina</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.17712v1-abstract-short" style="display: inline;"> In this work we study the problem of using machine-learned predictions to improve the performance of online algorithms. We consider two classical problems, ski rental and non-clairvoyant job scheduling, and obtain new online algorithms that use predictions to make their decisions. These algorithms are oblivious to the performance of the predictor, improve with better predictions, but do not degrad… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.17712v1-abstract-full').style.display = 'inline'; document.getElementById('2407.17712v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.17712v1-abstract-full" style="display: none;"> In this work we study the problem of using machine-learned predictions to improve the performance of online algorithms. We consider two classical problems, ski rental and non-clairvoyant job scheduling, and obtain new online algorithms that use predictions to make their decisions. These algorithms are oblivious to the performance of the predictor, improve with better predictions, but do not degrade much if the predictions are poor. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.17712v1-abstract-full').style.display = 'none'; document.getElementById('2407.17712v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Conference version appeared in Neurips 2018</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.15324">arXiv:2407.15324</a> <span> [<a href="https://arxiv.org/pdf/2407.15324">pdf</a>, <a href="https://arxiv.org/format/2407.15324">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Systems and Control">eess.SY</span> <span class="tag is-small is-grey 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> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Dynamical Systems">math.DS</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Optimization and Control">math.OC</span> </div> </div> <p class="title is-5 mathjax"> Cooperative Salvo Guidance over Leader-Follower Network with Free-Will Arbitrary Time Convergence </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Pal%2C+R+S">Rajib Shekhar Pal</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S+R">Shashi Ranjan Kumar</a>, <a href="/search/cs?searchtype=author&query=Mukherjee%2C+D">Dwaipayan Mukherjee</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.15324v1-abstract-short" style="display: inline;"> A cooperative salvo strategy is proposed in this paper which achieves consensus among the interceptors within a pre-defined arbitrary settling time. Considering non-linear engagement kinematics and a system lag to capture the effect of interceptor autopilot as present in realistic interception scenarios, the guidance schemes use the time-to-go estimates of the interceptors in order to achieve simu… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.15324v1-abstract-full').style.display = 'inline'; document.getElementById('2407.15324v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.15324v1-abstract-full" style="display: none;"> A cooperative salvo strategy is proposed in this paper which achieves consensus among the interceptors within a pre-defined arbitrary settling time. Considering non-linear engagement kinematics and a system lag to capture the effect of interceptor autopilot as present in realistic interception scenarios, the guidance schemes use the time-to-go estimates of the interceptors in order to achieve simultaneous interception of a stationary target at a pre-determined impact time. The guidance scheme ensures that consensus among the time-to-go estimates of the interceptors is achieved within a settling time whose upper bound can be pre-specified arbitrarily independent of the initial conditions or design parameters. The efficacy of the proposed guidance strategy is demonstrated using numerical simulations with varied conditions of initial position, velocities and heading angle errors of the interceptors as well as different desired impact times. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.15324v1-abstract-full').style.display = 'none'; document.getElementById('2407.15324v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.13833">arXiv:2407.13833</a> <span> [<a href="https://arxiv.org/pdf/2407.13833">pdf</a>, <a href="https://arxiv.org/format/2407.13833">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Phi-3 Safety Post-Training: Aligning Language Models with a "Break-Fix" Cycle </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Haider%2C+E">Emman Haider</a>, <a href="/search/cs?searchtype=author&query=Perez-Becker%2C+D">Daniel Perez-Becker</a>, <a href="/search/cs?searchtype=author&query=Portet%2C+T">Thomas Portet</a>, <a href="/search/cs?searchtype=author&query=Madan%2C+P">Piyush Madan</a>, <a href="/search/cs?searchtype=author&query=Garg%2C+A">Amit Garg</a>, <a href="/search/cs?searchtype=author&query=Ashfaq%2C+A">Atabak Ashfaq</a>, <a href="/search/cs?searchtype=author&query=Majercak%2C+D">David Majercak</a>, <a href="/search/cs?searchtype=author&query=Wen%2C+W">Wen Wen</a>, <a href="/search/cs?searchtype=author&query=Kim%2C+D">Dongwoo Kim</a>, <a href="/search/cs?searchtype=author&query=Yang%2C+Z">Ziyi Yang</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+J">Jianwen Zhang</a>, <a href="/search/cs?searchtype=author&query=Sharma%2C+H">Hiteshi Sharma</a>, <a href="/search/cs?searchtype=author&query=Bullwinkel%2C+B">Blake Bullwinkel</a>, <a href="/search/cs?searchtype=author&query=Pouliot%2C+M">Martin Pouliot</a>, <a href="/search/cs?searchtype=author&query=Minnich%2C+A">Amanda Minnich</a>, <a href="/search/cs?searchtype=author&query=Chawla%2C+S">Shiven Chawla</a>, <a href="/search/cs?searchtype=author&query=Herrera%2C+S">Solianna Herrera</a>, <a href="/search/cs?searchtype=author&query=Warreth%2C+S">Shahed Warreth</a>, <a href="/search/cs?searchtype=author&query=Engler%2C+M">Maggie Engler</a>, <a href="/search/cs?searchtype=author&query=Lopez%2C+G">Gary Lopez</a>, <a href="/search/cs?searchtype=author&query=Chikanov%2C+N">Nina Chikanov</a>, <a href="/search/cs?searchtype=author&query=Dheekonda%2C+R+S+R">Raja Sekhar Rao Dheekonda</a>, <a href="/search/cs?searchtype=author&query=Jagdagdorj%2C+B">Bolor-Erdene Jagdagdorj</a>, <a href="/search/cs?searchtype=author&query=Lutz%2C+R">Roman Lutz</a>, <a href="/search/cs?searchtype=author&query=Lundeen%2C+R">Richard Lundeen</a> , et al. (6 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="2407.13833v2-abstract-short" style="display: inline;"> Recent innovations in language model training have demonstrated that it is possible to create highly performant models that are small enough to run on a smartphone. As these models are deployed in an increasing number of domains, it is critical to ensure that they are aligned with human preferences and safety considerations. In this report, we present our methodology for safety aligning the Phi-3… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.13833v2-abstract-full').style.display = 'inline'; document.getElementById('2407.13833v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.13833v2-abstract-full" style="display: none;"> Recent innovations in language model training have demonstrated that it is possible to create highly performant models that are small enough to run on a smartphone. As these models are deployed in an increasing number of domains, it is critical to ensure that they are aligned with human preferences and safety considerations. In this report, we present our methodology for safety aligning the Phi-3 series of language models. We utilized a "break-fix" cycle, performing multiple rounds of dataset curation, safety post-training, benchmarking, red teaming, and vulnerability identification to cover a variety of harm areas in both single and multi-turn scenarios. Our results indicate that this approach iteratively improved the performance of the Phi-3 models across a wide range of responsible AI benchmarks. Finally, we include additional red teaming strategies and evaluations that were used to test the safety behavior of Phi-3.5-mini and Phi-3.5-MoE, which were optimized for multilingual capabilities. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.13833v2-abstract-full').style.display = 'none'; document.getElementById('2407.13833v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 22 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 18 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.13153">arXiv:2407.13153</a> <span> [<a href="https://arxiv.org/pdf/2407.13153">pdf</a>, <a href="https://arxiv.org/format/2407.13153">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="Cryptography and Security">cs.CR</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="Sound">cs.SD</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Audio and Speech Processing">eess.AS</span> </div> </div> <p class="title is-5 mathjax"> Preset-Voice Matching for Privacy Regulated Speech-to-Speech Translation Systems </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Platnick%2C+D">Daniel Platnick</a>, <a href="/search/cs?searchtype=author&query=Abdelnour%2C+B">Bishoy Abdelnour</a>, <a href="/search/cs?searchtype=author&query=Earl%2C+E">Eamon Earl</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rahul Kumar</a>, <a href="/search/cs?searchtype=author&query=Rezaei%2C+Z">Zahra Rezaei</a>, <a href="/search/cs?searchtype=author&query=Tsangaris%2C+T">Thomas Tsangaris</a>, <a href="/search/cs?searchtype=author&query=Lagum%2C+F">Faraj Lagum</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.13153v1-abstract-short" style="display: inline;"> In recent years, there has been increased demand for speech-to-speech translation (S2ST) systems in industry settings. Although successfully commercialized, cloning-based S2ST systems expose their distributors to liabilities when misused by individuals and can infringe on personality rights when exploited by media organizations. This work proposes a regulated S2ST framework called Preset-Voice Mat… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.13153v1-abstract-full').style.display = 'inline'; document.getElementById('2407.13153v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.13153v1-abstract-full" style="display: none;"> In recent years, there has been increased demand for speech-to-speech translation (S2ST) systems in industry settings. Although successfully commercialized, cloning-based S2ST systems expose their distributors to liabilities when misused by individuals and can infringe on personality rights when exploited by media organizations. This work proposes a regulated S2ST framework called Preset-Voice Matching (PVM). PVM removes cross-lingual voice cloning in S2ST by first matching the input voice to a similar prior consenting speaker voice in the target-language. With this separation, PVM avoids cloning the input speaker, ensuring PVM systems comply with regulations and reduce risk of misuse. Our results demonstrate PVM can significantly improve S2ST system run-time in multi-speaker settings and the naturalness of S2ST synthesized speech. To our knowledge, PVM is the first explicitly regulated S2ST framework leveraging similarly-matched preset-voices for dynamic S2ST tasks. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.13153v1-abstract-full').style.display = 'none'; document.getElementById('2407.13153v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 18 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted to the ACL PrivateNLP 2024 Workshop, 7 pages, 2 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/2407.12481">arXiv:2407.12481</a> <span> [<a href="https://arxiv.org/pdf/2407.12481">pdf</a>, <a href="https://arxiv.org/format/2407.12481">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"> Pretraining Data and Tokenizer for Indic LLM </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rahul Kumar</a>, <a href="/search/cs?searchtype=author&query=Kakde%2C+S">Shubham Kakde</a>, <a href="/search/cs?searchtype=author&query=Rajput%2C+D">Divyansh Rajput</a>, <a href="/search/cs?searchtype=author&query=Ibrahim%2C+D">Daud Ibrahim</a>, <a href="/search/cs?searchtype=author&query=Nahata%2C+R">Rishabh Nahata</a>, <a href="/search/cs?searchtype=author&query=Sowjanya%2C+P">Pidathala Sowjanya</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+D">Deepak Kumar</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.12481v1-abstract-short" style="display: inline;"> We present a novel approach to data preparation for developing multilingual Indic large language model. Our meticulous data acquisition spans open-source and proprietary sources, including Common Crawl, Indic books, news articles, and Wikipedia, ensuring a diverse and rich linguistic representation. For each Indic language, we design a custom preprocessing pipeline to effectively eliminate redunda… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.12481v1-abstract-full').style.display = 'inline'; document.getElementById('2407.12481v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.12481v1-abstract-full" style="display: none;"> We present a novel approach to data preparation for developing multilingual Indic large language model. Our meticulous data acquisition spans open-source and proprietary sources, including Common Crawl, Indic books, news articles, and Wikipedia, ensuring a diverse and rich linguistic representation. For each Indic language, we design a custom preprocessing pipeline to effectively eliminate redundant and low-quality text content. Additionally, we perform deduplication on Common Crawl data to address the redundancy present in 70% of the crawled web pages. This study focuses on developing high-quality data, optimizing tokenization for our multilingual dataset for Indic large language models with 3B and 7B parameters, engineered for superior performance in Indic languages. We introduce a novel multilingual tokenizer training strategy, demonstrating our custom-trained Indic tokenizer outperforms the state-of-the-art OpenAI Tiktoken tokenizer, achieving a superior token-to-word ratio for Indic languages. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.12481v1-abstract-full').style.display = 'none'; document.getElementById('2407.12481v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.10837">arXiv:2407.10837</a> <span> [<a href="https://arxiv.org/pdf/2407.10837">pdf</a>, <a href="https://arxiv.org/format/2407.10837">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Systems and Control">eess.SY</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Robotics">cs.RO</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Dynamical Systems">math.DS</span> </div> </div> <p class="title is-5 mathjax"> Trajectory Tracking for Unmanned Aerial Vehicles in 3D Spaces under Motion Constraints </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+S">Saurabh Kumar</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S+R">Shashi Ranjan Kumar</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Abhinav Sinha</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.10837v1-abstract-short" style="display: inline;"> This article presents a three-dimensional nonlinear trajectory tracking control strategy for unmanned aerial vehicles (UAVs) in the presence of spatial constraints. As opposed to many existing control strategies, which do not consider spatial constraints, the proposed strategy considers spatial constraints on each degree of freedom movement of the UAV. Such consideration makes the design appealing… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.10837v1-abstract-full').style.display = 'inline'; document.getElementById('2407.10837v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.10837v1-abstract-full" style="display: none;"> This article presents a three-dimensional nonlinear trajectory tracking control strategy for unmanned aerial vehicles (UAVs) in the presence of spatial constraints. As opposed to many existing control strategies, which do not consider spatial constraints, the proposed strategy considers spatial constraints on each degree of freedom movement of the UAV. Such consideration makes the design appealing for many practical applications, such as pipeline inspection, boundary tracking, etc. The proposed design accounts for the limited information about the inertia matrix, thereby affirming its inherent robustness against unmodeled dynamics and other imperfections. We rigorously show that the UAV will converge to its desired path by maintaining bounded position, orientation, and linear and angular speeds. Finally, we demonstrate the effectiveness of the proposed strategy through various numerical simulations. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.10837v1-abstract-full').style.display = 'none'; document.getElementById('2407.10837v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.08655">arXiv:2407.08655</a> <span> [<a href="https://arxiv.org/pdf/2407.08655">pdf</a>, <a href="https://arxiv.org/format/2407.08655">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Image and Video Processing">eess.IV</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Medical Physics">physics.med-ph</span> </div> </div> <p class="title is-5 mathjax"> SPOCKMIP: Segmentation of Vessels in MRAs with Enhanced Continuity using Maximum Intensity Projection as Loss </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Radhakrishna%2C+C">Chethan Radhakrishna</a>, <a href="/search/cs?searchtype=author&query=Chintalapati%2C+K+V">Karthikesh Varma Chintalapati</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+S+C+H+R">Sri Chandana Hudukula Ram Kumar</a>, <a href="/search/cs?searchtype=author&query=Sutrave%2C+R">Raviteja Sutrave</a>, <a href="/search/cs?searchtype=author&query=Mattern%2C+H">Hendrik Mattern</a>, <a href="/search/cs?searchtype=author&query=Speck%2C+O">Oliver Speck</a>, <a href="/search/cs?searchtype=author&query=N%C3%BCrnberger%2C+A">Andreas N眉rnberger</a>, <a href="/search/cs?searchtype=author&query=Chatterjee%2C+S">Soumick Chatterjee</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.08655v1-abstract-short" style="display: inline;"> Identification of vessel structures of different sizes in biomedical images is crucial in the diagnosis of many neurodegenerative diseases. However, the sparsity of good-quality annotations of such images makes the task of vessel segmentation challenging. Deep learning offers an efficient way to segment vessels of different sizes by learning their high-level feature representations and the spatial… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.08655v1-abstract-full').style.display = 'inline'; document.getElementById('2407.08655v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.08655v1-abstract-full" style="display: none;"> Identification of vessel structures of different sizes in biomedical images is crucial in the diagnosis of many neurodegenerative diseases. However, the sparsity of good-quality annotations of such images makes the task of vessel segmentation challenging. Deep learning offers an efficient way to segment vessels of different sizes by learning their high-level feature representations and the spatial continuity of such features across dimensions. Semi-supervised patch-based approaches have been effective in identifying small vessels of one to two voxels in diameter. This study focuses on improving the segmentation quality by considering the spatial correlation of the features using the Maximum Intensity Projection~(MIP) as an additional loss criterion. Two methods are proposed with the incorporation of MIPs of label segmentation on the single~(z-axis) and multiple perceivable axes of the 3D volume. The proposed MIP-based methods produce segmentations with improved vessel continuity, which is evident in visual examinations of ROIs. Patch-based training is improved by introducing an additional loss term, MIP loss, to penalise the predicted discontinuity of vessels. A training set of 14 volumes is selected from the StudyForrest dataset comprising of 18 7-Tesla 3D Time-of-Flight~(ToF) Magnetic Resonance Angiography (MRA) images. The generalisation performance of the method is evaluated using the other unseen volumes in the dataset. It is observed that the proposed method with multi-axes MIP loss produces better quality segmentations with a median Dice of $80.245 \pm 0.129$. Also, the method with single-axis MIP loss produces segmentations with a median Dice of $79.749 \pm 0.109$. Furthermore, a visual comparison of the ROIs in the predicted segmentation reveals a significant improvement in the continuity of the vessels when MIP loss is incorporated into training. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.08655v1-abstract-full').style.display = 'none'; document.getElementById('2407.08655v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.08041">arXiv:2407.08041</a> <span> [<a href="https://arxiv.org/pdf/2407.08041">pdf</a>, <a href="https://arxiv.org/format/2407.08041">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"> TACLE: Task and Class-aware Exemplar-free Semi-supervised Class Incremental Learning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kalla%2C+J">Jayateja Kalla</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rohit Kumar</a>, <a href="/search/cs?searchtype=author&query=Biswas%2C+S">Soma Biswas</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.08041v1-abstract-short" style="display: inline;"> We propose a novel TACLE (TAsk and CLass-awarE) framework to address the relatively unexplored and challenging problem of exemplar-free semi-supervised class incremental learning. In this scenario, at each new task, the model has to learn new classes from both (few) labeled and unlabeled data without access to exemplars from previous classes. In addition to leveraging the capabilities of pre-train… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.08041v1-abstract-full').style.display = 'inline'; document.getElementById('2407.08041v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.08041v1-abstract-full" style="display: none;"> We propose a novel TACLE (TAsk and CLass-awarE) framework to address the relatively unexplored and challenging problem of exemplar-free semi-supervised class incremental learning. In this scenario, at each new task, the model has to learn new classes from both (few) labeled and unlabeled data without access to exemplars from previous classes. In addition to leveraging the capabilities of pre-trained models, TACLE proposes a novel task-adaptive threshold, thereby maximizing the utilization of the available unlabeled data as incremental learning progresses. Additionally, to enhance the performance of the under-represented classes within each task, we propose a class-aware weighted cross-entropy loss. We also exploit the unlabeled data for classifier alignment, which further enhances the model performance. Extensive experiments on benchmark datasets, namely CIFAR10, CIFAR100, and ImageNet-Subset100 demonstrate the effectiveness of the proposed TACLE framework. We further showcase its effectiveness when the unlabeled data is imbalanced and also for the extreme case of one labeled example per class. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.08041v1-abstract-full').style.display = 'none'; document.getElementById('2407.08041v1-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 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.07786">arXiv:2407.07786</a> <span> [<a href="https://arxiv.org/pdf/2407.07786">pdf</a>, <a href="https://arxiv.org/format/2407.07786">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> </div> <div class="is-inline-block" style="margin-left: 0.5rem"> <div class="tags has-addons"> <span class="tag is-dark is-size-7">doi</span> <span class="tag is-light is-size-7"><a class="" href="https://doi.org/10.1145/3678884.3687147">10.1145/3678884.3687147 <i class="fa fa-external-link" aria-hidden="true"></i></a></span> </div> </div> </div> <p class="title is-5 mathjax"> The Human Factor in AI Red Teaming: Perspectives from Social and Collaborative Computing </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Zhang%2C+A+Q">Alice Qian Zhang</a>, <a href="/search/cs?searchtype=author&query=Shaw%2C+R">Ryland Shaw</a>, <a href="/search/cs?searchtype=author&query=Anthis%2C+J+R">Jacy Reese Anthis</a>, <a href="/search/cs?searchtype=author&query=Milton%2C+A">Ashlee Milton</a>, <a href="/search/cs?searchtype=author&query=Tseng%2C+E">Emily Tseng</a>, <a href="/search/cs?searchtype=author&query=Suh%2C+J">Jina Suh</a>, <a href="/search/cs?searchtype=author&query=Ahmad%2C+L">Lama Ahmad</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R+S+S">Ram Shankar Siva Kumar</a>, <a href="/search/cs?searchtype=author&query=Posada%2C+J">Julian Posada</a>, <a href="/search/cs?searchtype=author&query=Shestakofsky%2C+B">Benjamin Shestakofsky</a>, <a href="/search/cs?searchtype=author&query=Roberts%2C+S+T">Sarah T. Roberts</a>, <a href="/search/cs?searchtype=author&query=Gray%2C+M+L">Mary L. Gray</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.07786v2-abstract-short" style="display: inline;"> Rapid progress in general-purpose AI has sparked significant interest in "red teaming," a practice of adversarial testing originating in military and cybersecurity applications. AI red teaming raises many questions about the human factor, such as how red teamers are selected, biases and blindspots in how tests are conducted, and harmful content's psychological effects on red teamers. A growing bod… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.07786v2-abstract-full').style.display = 'inline'; document.getElementById('2407.07786v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.07786v2-abstract-full" style="display: none;"> Rapid progress in general-purpose AI has sparked significant interest in "red teaming," a practice of adversarial testing originating in military and cybersecurity applications. AI red teaming raises many questions about the human factor, such as how red teamers are selected, biases and blindspots in how tests are conducted, and harmful content's psychological effects on red teamers. A growing body of HCI and CSCW literature examines related practices-including data labeling, content moderation, and algorithmic auditing. However, few, if any have investigated red teaming itself. Future studies may explore topics ranging from fairness to mental health and other areas of potential harm. We aim to facilitate a community of researchers and practitioners who can begin to meet these challenges with creativity, innovation, and thoughtful reflection. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.07786v2-abstract-full').style.display = 'none'; document.getElementById('2407.07786v2-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 September, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 10 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Updated with camera-ready version</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2407.06093">arXiv:2407.06093</a> <span> [<a href="https://arxiv.org/pdf/2407.06093">pdf</a>, <a href="https://arxiv.org/format/2407.06093">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> </div> </div> <p class="title is-5 mathjax"> Artificial Intuition: Efficient Classification of Scientific Abstracts </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Sakhrani%2C+H">Harsh Sakhrani</a>, <a href="/search/cs?searchtype=author&query=Pervez%2C+N">Naseela Pervez</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+A+R">Anirudh Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Morstatter%2C+F">Fred Morstatter</a>, <a href="/search/cs?searchtype=author&query=Reed%2C+A+G">Alexandra Graddy Reed</a>, <a href="/search/cs?searchtype=author&query=Belz%2C+A">Andrea Belz</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2407.06093v1-abstract-short" style="display: inline;"> It is desirable to coarsely classify short scientific texts, such as grant or publication abstracts, for strategic insight or research portfolio management. These texts efficiently transmit dense information to experts possessing a rich body of knowledge to aid interpretation. Yet this task is remarkably difficult to automate because of brevity and the absence of context. To address this gap, we h… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.06093v1-abstract-full').style.display = 'inline'; document.getElementById('2407.06093v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2407.06093v1-abstract-full" style="display: none;"> It is desirable to coarsely classify short scientific texts, such as grant or publication abstracts, for strategic insight or research portfolio management. These texts efficiently transmit dense information to experts possessing a rich body of knowledge to aid interpretation. Yet this task is remarkably difficult to automate because of brevity and the absence of context. To address this gap, we have developed a novel approach to generate and appropriately assign coarse domain-specific labels. We show that a Large Language Model (LLM) can provide metadata essential to the task, in a process akin to the augmentation of supplemental knowledge representing human intuition, and propose a workflow. As a pilot study, we use a corpus of award abstracts from the National Aeronautics and Space Administration (NASA). We develop new assessment tools in concert with established performance metrics. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2407.06093v1-abstract-full').style.display = 'none'; document.getElementById('2407.06093v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 8 July, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> July 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.19040">arXiv:2406.19040</a> <span> [<a href="https://arxiv.org/pdf/2406.19040">pdf</a>, <a href="https://arxiv.org/ps/2406.19040">ps</a>, <a href="https://arxiv.org/format/2406.19040">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="Cryptography and Security">cs.CR</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Data Structures and Algorithms">cs.DS</span> </div> </div> <p class="title is-5 mathjax"> On Convex Optimization with Semi-Sensitive Features </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Meka%2C+R">Raghu Meka</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan 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="2406.19040v1-abstract-short" style="display: inline;"> We study the differentially private (DP) empirical risk minimization (ERM) problem under the semi-sensitive DP setting where only some features are sensitive. This generalizes the Label DP setting where only the label is sensitive. We give improved upper and lower bounds on the excess risk for DP-ERM. In particular, we show that the error only scales polylogarithmically in terms of the sensitive d… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.19040v1-abstract-full').style.display = 'inline'; document.getElementById('2406.19040v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.19040v1-abstract-full" style="display: none;"> We study the differentially private (DP) empirical risk minimization (ERM) problem under the semi-sensitive DP setting where only some features are sensitive. This generalizes the Label DP setting where only the label is sensitive. We give improved upper and lower bounds on the excess risk for DP-ERM. In particular, we show that the error only scales polylogarithmically in terms of the sensitive domain size, improving upon previous results that scale polynomially in the sensitive domain size (Ghazi et al., 2021). <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.19040v1-abstract-full').style.display = 'none'; document.getElementById('2406.19040v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 27 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">To appear in COLT 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.16305">arXiv:2406.16305</a> <span> [<a href="https://arxiv.org/pdf/2406.16305">pdf</a>, <a href="https://arxiv.org/ps/2406.16305">ps</a>, <a href="https://arxiv.org/format/2406.16305">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Data Structures and Algorithms">cs.DS</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Cryptography and Security">cs.CR</span> </div> </div> <p class="title is-5 mathjax"> On Computing Pairwise Statistics with Local Differential Privacy </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Sealfon%2C+A">Adam Sealfon</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2406.16305v1-abstract-short" style="display: inline;"> We study the problem of computing pairwise statistics, i.e., ones of the form $\binom{n}{2}^{-1} \sum_{i \ne j} f(x_i, x_j)$, where $x_i$ denotes the input to the $i$th user, with differential privacy (DP) in the local model. This formulation captures important metrics such as Kendall's $蟿$ coefficient, Area Under Curve, Gini's mean difference, Gini's entropy, etc. We give several novel and generi… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.16305v1-abstract-full').style.display = 'inline'; document.getElementById('2406.16305v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.16305v1-abstract-full" style="display: none;"> We study the problem of computing pairwise statistics, i.e., ones of the form $\binom{n}{2}^{-1} \sum_{i \ne j} f(x_i, x_j)$, where $x_i$ denotes the input to the $i$th user, with differential privacy (DP) in the local model. This formulation captures important metrics such as Kendall's $蟿$ coefficient, Area Under Curve, Gini's mean difference, Gini's entropy, etc. We give several novel and generic algorithms for the problem, leveraging techniques from DP algorithms for linear queries. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.16305v1-abstract-full').style.display = 'none'; document.getElementById('2406.16305v1-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 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Published in NeurIPS 2023</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.16135">arXiv:2406.16135</a> <span> [<a href="https://arxiv.org/pdf/2406.16135">pdf</a>, <a href="https://arxiv.org/format/2406.16135">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Machine Learning">cs.LG</span> </div> </div> <p class="title is-5 mathjax"> Crosslingual Capabilities and Knowledge Barriers in Multilingual Large Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chua%2C+L">Lynn Chua</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yangsibo Huang</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Amer Sinha</a>, <a href="/search/cs?searchtype=author&query=Xie%2C+C">Chulin Xie</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan 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="2406.16135v1-abstract-short" style="display: inline;"> Large language models (LLMs) are typically multilingual due to pretraining on diverse multilingual corpora. But can these models relate corresponding concepts across languages, effectively being crosslingual? This study evaluates six state-of-the-art LLMs on inherently crosslingual tasks. We observe that while these models show promising surface-level crosslingual abilities on machine translation… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.16135v1-abstract-full').style.display = 'inline'; document.getElementById('2406.16135v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.16135v1-abstract-full" style="display: none;"> Large language models (LLMs) are typically multilingual due to pretraining on diverse multilingual corpora. But can these models relate corresponding concepts across languages, effectively being crosslingual? This study evaluates six state-of-the-art LLMs on inherently crosslingual tasks. We observe that while these models show promising surface-level crosslingual abilities on machine translation and embedding space analyses, they struggle with deeper crosslingual knowledge transfer, revealing a crosslingual knowledge barrier in both general (MMLU benchmark) and domain-specific (Harry Potter quiz) contexts. We observe that simple inference-time mitigation methods offer only limited improvement. On the other hand, we propose fine-tuning of LLMs on mixed-language data, which effectively reduces these gaps, even when using out-of-domain datasets like WikiText. Our findings suggest the need for explicit optimization to unlock the full crosslingual potential of LLMs. Our code is publicly available at https://github.com/google-research/crosslingual-knowledge-barriers. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.16135v1-abstract-full').style.display = 'none'; document.getElementById('2406.16135v1-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 23 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.15470">arXiv:2406.15470</a> <span> [<a href="https://arxiv.org/pdf/2406.15470">pdf</a>, <a href="https://arxiv.org/format/2406.15470">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="Social and Information Networks">cs.SI</span> </div> </div> <p class="title is-5 mathjax"> Mental Disorder Classification via Temporal Representation of Text </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Raja Kumar</a>, <a href="/search/cs?searchtype=author&query=Maharaj%2C+K">Kishan Maharaj</a>, <a href="/search/cs?searchtype=author&query=Saxena%2C+A">Ashita Saxena</a>, <a href="/search/cs?searchtype=author&query=Bhattacharyya%2C+P">Pushpak Bhattacharyya</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2406.15470v2-abstract-short" style="display: inline;"> Mental disorders pose a global challenge, aggravated by the shortage of qualified mental health professionals. Mental disorder prediction from social media posts by current LLMs is challenging due to the complexities of sequential text data and the limited context length of language models. Current language model-based approaches split a single data instance into multiple chunks to compensate for… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15470v2-abstract-full').style.display = 'inline'; document.getElementById('2406.15470v2-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.15470v2-abstract-full" style="display: none;"> Mental disorders pose a global challenge, aggravated by the shortage of qualified mental health professionals. Mental disorder prediction from social media posts by current LLMs is challenging due to the complexities of sequential text data and the limited context length of language models. Current language model-based approaches split a single data instance into multiple chunks to compensate for limited context size. The predictive model is then applied to each chunk individually, and the most voted output is selected as the final prediction. This results in the loss of inter-post dependencies and important time variant information, leading to poor performance. We propose a novel framework which first compresses the large sequence of chronologically ordered social media posts into a series of numbers. We then use this time variant representation for mental disorder classification. We demonstrate the generalization capabilities of our framework by outperforming the current SOTA in three different mental conditions: depression, self-harm, and anorexia, with an absolute improvement of 5% in the F1 score. We investigate the situation where current data instances fall within the context length of language models and present empirical results highlighting the importance of temporal properties of textual data. Furthermore, we utilize the proposed framework for a cross-domain study, exploring commonalities across disorders and the possibility of inter-domain data usage. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15470v2-abstract-full').style.display = 'none'; document.getElementById('2406.15470v2-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 6 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 15 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">RK and KM contributed equally to this work, 15 pages, 5 figures, 9 table</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.15335">arXiv:2406.15335</a> <span> [<a href="https://arxiv.org/pdf/2406.15335">pdf</a>, <a href="https://arxiv.org/format/2406.15335">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="Computers and Society">cs.CY</span> </div> </div> <p class="title is-5 mathjax"> Keystroke Dynamics Against Academic Dishonesty in the Age of LLMs </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Kundu%2C+D">Debnath Kundu</a>, <a href="/search/cs?searchtype=author&query=Mehta%2C+A">Atharva Mehta</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Rajesh Kumar</a>, <a href="/search/cs?searchtype=author&query=Lal%2C+N">Naman Lal</a>, <a href="/search/cs?searchtype=author&query=Anand%2C+A">Avinash Anand</a>, <a href="/search/cs?searchtype=author&query=Singh%2C+A">Apoorv Singh</a>, <a href="/search/cs?searchtype=author&query=Shah%2C+R+R">Rajiv Ratn Shah</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2406.15335v1-abstract-short" style="display: inline;"> The transition to online examinations and assignments raises significant concerns about academic integrity. Traditional plagiarism detection systems often struggle to identify instances of intelligent cheating, particularly when students utilize advanced generative AI tools to craft their responses. This study proposes a keystroke dynamics-based method to differentiate between bona fide and assist… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15335v1-abstract-full').style.display = 'inline'; document.getElementById('2406.15335v1-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.15335v1-abstract-full" style="display: none;"> The transition to online examinations and assignments raises significant concerns about academic integrity. Traditional plagiarism detection systems often struggle to identify instances of intelligent cheating, particularly when students utilize advanced generative AI tools to craft their responses. This study proposes a keystroke dynamics-based method to differentiate between bona fide and assisted writing within academic contexts. To facilitate this, a dataset was developed to capture the keystroke patterns of individuals engaged in writing tasks, both with and without the assistance of generative AI. The detector, trained using a modified TypeNet architecture, achieved accuracies ranging from 74.98% to 85.72% in condition-specific scenarios and from 52.24% to 80.54% in condition-agnostic scenarios. The findings highlight significant differences in keystroke dynamics between genuine and assisted writing. The outcomes of this study enhance our understanding of how users interact with generative AI and have implications for improving the reliability of digital educational platforms. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.15335v1-abstract-full').style.display = 'none'; document.getElementById('2406.15335v1-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 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Accepted for publication at The IEEE International Joint Conference on Biometrics (IJCB2024), contains 9 pages, 3 figures, 3 tables</span> </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">ACM Class:</span> 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/2406.14322">arXiv:2406.14322</a> <span> [<a href="https://arxiv.org/pdf/2406.14322">pdf</a>, <a href="https://arxiv.org/format/2406.14322">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="Cryptography and Security">cs.CR</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"> Mind the Privacy Unit! User-Level Differential Privacy for Language Model Fine-Tuning </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Chua%2C+L">Lynn Chua</a>, <a href="/search/cs?searchtype=author&query=Ghazi%2C+B">Badih Ghazi</a>, <a href="/search/cs?searchtype=author&query=Huang%2C+Y">Yangsibo Huang</a>, <a href="/search/cs?searchtype=author&query=Kamath%2C+P">Pritish Kamath</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravi Kumar</a>, <a href="/search/cs?searchtype=author&query=Liu%2C+D">Daogao Liu</a>, <a href="/search/cs?searchtype=author&query=Manurangsi%2C+P">Pasin Manurangsi</a>, <a href="/search/cs?searchtype=author&query=Sinha%2C+A">Amer Sinha</a>, <a href="/search/cs?searchtype=author&query=Zhang%2C+C">Chiyuan 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="2406.14322v3-abstract-short" style="display: inline;"> Large language models (LLMs) have emerged as powerful tools for tackling complex tasks across diverse domains, but they also raise privacy concerns when fine-tuned on sensitive data due to potential memorization. While differential privacy (DP) offers a promising solution by ensuring models are 'almost indistinguishable' with or without any particular privacy unit, current evaluations on LLMs most… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.14322v3-abstract-full').style.display = 'inline'; document.getElementById('2406.14322v3-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.14322v3-abstract-full" style="display: none;"> Large language models (LLMs) have emerged as powerful tools for tackling complex tasks across diverse domains, but they also raise privacy concerns when fine-tuned on sensitive data due to potential memorization. While differential privacy (DP) offers a promising solution by ensuring models are 'almost indistinguishable' with or without any particular privacy unit, current evaluations on LLMs mostly treat each example (text record) as the privacy unit. This leads to uneven user privacy guarantees when contributions per user vary. We therefore study user-level DP motivated by applications where it necessary to ensure uniform privacy protection across users. We present a systematic evaluation of user-level DP for LLM fine-tuning on natural language generation tasks. Focusing on two mechanisms for achieving user-level DP guarantees, Group Privacy and User-wise DP-SGD, we investigate design choices like data selection strategies and parameter tuning for the best privacy-utility tradeoff. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.14322v3-abstract-full').style.display = 'none'; document.getElementById('2406.14322v3-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 16 August, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 20 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">Published as a conference paper at COLM 2024</span> </p> </li> <li class="arxiv-result"> <div class="is-marginless"> <p class="list-title is-inline-block"><a href="https://arxiv.org/abs/2406.11757">arXiv:2406.11757</a> <span> [<a href="https://arxiv.org/pdf/2406.11757">pdf</a>, <a href="https://arxiv.org/format/2406.11757">other</a>] </span> </p> <div class="tags is-inline-block"> <span class="tag is-small is-link tooltip is-tooltip-top" data-tooltip="Artificial Intelligence">cs.AI</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computation and Language">cs.CL</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Computers and Society">cs.CY</span> <span class="tag is-small is-grey tooltip is-tooltip-top" data-tooltip="Human-Computer Interaction">cs.HC</span> </div> </div> <p class="title is-5 mathjax"> STAR: SocioTechnical Approach to Red Teaming Language Models </p> <p class="authors"> <span class="search-hit">Authors:</span> <a href="/search/cs?searchtype=author&query=Weidinger%2C+L">Laura Weidinger</a>, <a href="/search/cs?searchtype=author&query=Mellor%2C+J">John Mellor</a>, <a href="/search/cs?searchtype=author&query=Pegueroles%2C+B+G">Bernat Guillen Pegueroles</a>, <a href="/search/cs?searchtype=author&query=Marchal%2C+N">Nahema Marchal</a>, <a href="/search/cs?searchtype=author&query=Kumar%2C+R">Ravin Kumar</a>, <a href="/search/cs?searchtype=author&query=Lum%2C+K">Kristian Lum</a>, <a href="/search/cs?searchtype=author&query=Akbulut%2C+C">Canfer Akbulut</a>, <a href="/search/cs?searchtype=author&query=Diaz%2C+M">Mark Diaz</a>, <a href="/search/cs?searchtype=author&query=Bergman%2C+S">Stevie Bergman</a>, <a href="/search/cs?searchtype=author&query=Rodriguez%2C+M">Mikel Rodriguez</a>, <a href="/search/cs?searchtype=author&query=Rieser%2C+V">Verena Rieser</a>, <a href="/search/cs?searchtype=author&query=Isaac%2C+W">William Isaac</a> </p> <p class="abstract mathjax"> <span class="has-text-black-bis has-text-weight-semibold">Abstract</span>: <span class="abstract-short has-text-grey-dark mathjax" id="2406.11757v4-abstract-short" style="display: inline;"> This research introduces STAR, a sociotechnical framework that improves on current best practices for red teaming safety of large language models. STAR makes two key contributions: it enhances steerability by generating parameterised instructions for human red teamers, leading to improved coverage of the risk surface. Parameterised instructions also provide more detailed insights into model failur… <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.11757v4-abstract-full').style.display = 'inline'; document.getElementById('2406.11757v4-abstract-short').style.display = 'none';">▽ More</a> </span> <span class="abstract-full has-text-grey-dark mathjax" id="2406.11757v4-abstract-full" style="display: none;"> This research introduces STAR, a sociotechnical framework that improves on current best practices for red teaming safety of large language models. STAR makes two key contributions: it enhances steerability by generating parameterised instructions for human red teamers, leading to improved coverage of the risk surface. Parameterised instructions also provide more detailed insights into model failures at no increased cost. Second, STAR improves signal quality by matching demographics to assess harms for specific groups, resulting in more sensitive annotations. STAR further employs a novel step of arbitration to leverage diverse viewpoints and improve label reliability, treating disagreement not as noise but as a valuable contribution to signal quality. <a class="is-size-7" style="white-space: nowrap;" onclick="document.getElementById('2406.11757v4-abstract-full').style.display = 'none'; document.getElementById('2406.11757v4-abstract-short').style.display = 'inline';">△ Less</a> </span> </p> <p class="is-size-7"><span class="has-text-black-bis has-text-weight-semibold">Submitted</span> 23 October, 2024; <span class="has-text-black-bis has-text-weight-semibold">v1</span> submitted 17 June, 2024; <span class="has-text-black-bis has-text-weight-semibold">originally announced</span> June 2024. </p> <p class="comments is-size-7"> <span class="has-text-black-bis has-text-weight-semibold">Comments:</span> <span class="has-text-grey-dark mathjax">8 pages, 5 figures, 5 pages appendix. * denotes equal contribution</span> </p> </li> </ol> <nav class="pagination is-small is-centered breathe-horizontal" role="navigation" aria-label="pagination"> <a href="" class="pagination-previous is-invisible">Previous </a> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=50" class="pagination-next" >Next </a> <ul class="pagination-list"> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=0" class="pagination-link is-current" aria-label="Goto page 1">1 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=50" class="pagination-link " aria-label="Page 2" aria-current="page">2 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=100" class="pagination-link " aria-label="Page 3" aria-current="page">3 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%2C+R&start=150" class="pagination-link " aria-label="Page 4" aria-current="page">4 </a> </li> <li> <a href="/search/?searchtype=author&query=Kumar%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