Найден корректный robots.txt
Файл управления сканированием сайта роботами
?
Robots.txt указывает поисковым роботам, какие страницы сканировать, а какие — нет. Ошибки в файле могут случайно закрыть важные разделы от индексации.
Robots.txt настроен корректно. Размер файла: 15056 байт. Загружен за: 0сек.
Проверяемая страница не запрещена в robots.txt.
Цепочка редиректов для файла robots.txt:
http://keremdayi.com/robots.txt
301 MovedPermanently
https://keremdayi.com/robots.txt
200 OK
Показать содержимое robots.txt
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>Arif Kerem Dayı</title>
<meta name="description"
content="Arif Kerem Dayı — PhD student at MIT CSAIL" />
<link rel="preconnect" href="https://fonts.googleapis.com" />
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin />
<link
href="https://fonts.googleapis.com/css2?family=Ubuntu:ital,wght@0,300;0,400;0,500;0,700;1,400&family=Ubuntu+Mono:wght@400;700&display=swap"
rel="stylesheet" />
<link rel="stylesheet" href="css/style.css" />
</head>
<body>
<!-- ===================== TOP NAVIGATION ===================== -->
<header class="nav">
<div class="nav__inner">
<a class="nav__brand" href="#top">
<span class="nav__mark">AKD</span>
<span class="nav__name">Arif Kerem Dayı</span>
</a>
<nav class="nav__links" aria-label="Primary">
<a href="#about" data-nav>About</a>
<a href="#publications" data-nav>Publications</a>
<a href="#experience" data-nav>Experience</a>
<a href="#teaching" data-nav>Teaching</a>
<a href="#writing" data-nav>Writing</a>
<!-- <a href="files/cv.pdf" class="nav__cv">CV ↗</a> -->
</nav>
</div>
</header>
<span id="top"></span>
<!-- ===================== PAGE LAYOUT ===================== -->
<div class="layout">
<!-- ---------- SIDEBAR ---------- -->
<aside class="sidebar">
<div class="card">
<img class="portrait" src="files/photo.jpeg"
alt="Portrait of Arif Kerem Dayı" />
<h1 class="sidebar__name">Arif Kerem Dayı</h1>
<p class="sidebar__role">PhD student, <span>MIT CSAIL</span></p>
<!-- <p class="sidebar__tag">Theory of diffusion models & sampling in -->
<!-- sequence models.</p> -->
<ul class="contact">
<li>
<a href="/cdn-cgi/l/email-protection#a1cac5c0d8c8e1ccc8d58fc4c5d4">
<svg viewBox="0 0 24 24" aria-hidden="true"><path d="M3 5h18a1 1 0 0 1 1 1v12a1 1 0 0 1-1 1H3a1 1 0 0 1-1-1V6a1 1 0 0 1 1-1Zm.4 2L12 12.6 20.6 7H3.4ZM20 8.9l-8 5.2-8-5.2V17h16V8.9Z"/></svg>
<span><span class="__cf_email__" data-cfemail="771c13160e1e371a1e0359121302">[email protected]</span></span>
</a>
</li>
<li>
<a href="https://scholar.google.com/citations?user=HfiPj7oAAAAJ&hl=en" target="_blank" rel="noopener noreferrer">
<svg viewBox="0 0 24 24" aria-hidden="true"><path d="M12 2 1 8l11 6 9-4.9V16h2V8L12 2ZM5 13.2V17c0 1.7 3.1 3 7 3s7-1.3 7-3v-3.8l-7 3.8-7-3.8Z"/></svg>
<span>Google Scholar</span>
</a>
</li>
<li>
<a href="https://github.com/keremdayi" target="_blank" rel="noopener noreferrer">
<svg viewBox="0 0 24 24" aria-hidden="true"><path d="M12 2C6.5 2 2 6.6 2 12.3c0 4.5 2.9 8.3 6.8 9.7.5.1.7-.2.7-.5v-1.7c-2.8.6-3.4-1.4-3.4-1.4-.4-1.2-1.1-1.5-1.1-1.5-.9-.6.1-.6.1-.6 1 .1 1.5 1 1.5 1 .9 1.6 2.4 1.1 3 .9.1-.7.4-1.1.6-1.4-2.2-.3-4.6-1.1-4.6-5 0-1.1.4-2 1-2.7-.1-.3-.4-1.3.1-2.7 0 0 .8-.3 2.7 1a9.3 9.3 0 0 1 5 0c1.9-1.3 2.7-1 2.7-1 .5 1.4.2 2.4.1 2.7.6.7 1 1.6 1 2.7 0 3.9-2.3 4.7-4.6 5 .4.3.7.9.7 1.9v2.8c0 .3.2.6.7.5 3.9-1.4 6.8-5.2 6.8-9.7C22 6.6 17.5 2 12 2Z"/></svg>
<span>GitHub</span>
</a>
</li>
<li>
<a href="files/cv.pdf" target="_blank" rel="noopener noreferrer">
<svg viewBox="0 0 24 24" aria-hidden="true"><path d="M6 2h8l4 4v16H6V2Zm7 1.5V7h3.5L13 3.5ZM8 12h8v1.5H8V12Zm0 3.5h8V17H8v-1.5ZM8 19h5v1.5H8V19Z"/></svg>
<span>Curriculum Vitae</span>
</a>
</li>
</ul>
</div>
</aside>
<!-- ---------- MAIN CONTENT ---------- -->
<main class="content">
<section id="about" class="section">
<h2 class="section__title"><span class="section__no">01</span>About</h2>
<div class="prose">
<p>I am a first-year <strong>Ph.D. student at MIT CSAIL</strong>, advised by
Profs.
<a href="https://people.csail.mit.edu/costis/" target="_blank" rel="noopener noreferrer">Constantinos
Daskalakis</a> and
<a href="https://locomotion.csail.mit.edu/russt.html" target="_blank" rel="noopener noreferrer">Russ
Tedrake</a>. My research interests lie broadly in the theory of machine learning. More specifically, I am interested in the theory of post training in the context of diffusion models (and their applications to robotics), continual learning in language models, and other questions related to generative modeling. Previously, I worked on various questions related to the theory of low-rank fine tuning in neural networks, distributed optimization and learning in multiagent systems, and strategic interactions in LLMs.
</p>
<p>
During summer 2026, I am a research intern at <strong>Microsoft Research New England</strong>, advised by <a href="https://dylanfoster.net/" target="_blank" rel="noopener noreferrer">Dylan Foster</a>.
</p>
<p>Previously, I received an <strong>A.B. in Computer Science and
Mathematics</strong> and an <strong>S.M. in Computer Science</strong>
from <strong>Harvard University</strong> in May 2025. During my
undergraduate education, I was fortunate to work on the theory of
low-rank fine-tuning with SGD (advised by Prof.
<a href="https://sitanchen.com" target="_blank" rel="noopener noreferrer">Sitan Chen</a>), and
various problems in distributed / multi-agent optimization (advised by
Profs.
<a href="https://react.seas.harvard.edu/people/stephanie-gil" target="_blank" rel="noopener noreferrer">Stephanie
Gil</a> and
<a href="https://faculty.engineering.asu.edu/nedich" target="_blank" rel="noopener noreferrer">Angelia
Nedich</a>).</p>
</div>
</section>
<section id="publications" class="section">
<h2 class="section__title"><span class="section__no">02</span>Publications</h2>
<p class="section__lede">A selection of work below. For the full list, see my
<a href="https://scholar.google.com/citations?user=HfiPj7oAAAAJ&hl=en" target="_blank" rel="noopener noreferrer">Google
Scholar</a> page.</p>
<ol class="pubs">
<li class="pub">
<a class="pub__title" href="https://scholar.google.com/citations?view_op=view_citation&hl=en&user=HfiPj7oAAAAJ&citation_for_view=HfiPj7oAAAAJ:WF5omc3nYNoC"
target="_blank" rel="noopener noreferrer">Low-rank fine-tuning lies between lazy training and feature
learning</a>
<p class="pub__authors"><strong>Arif Kerem Dayı</strong>, Sitan Chen</p>
<p class="pub__venue">Conference on Learning Theory (COLT) <span>2025</span></p>
</li>
<li class="pub">
<a class="pub__title" href="https://scholar.google.com/citations?view_op=view_citation&hl=en&user=HfiPj7oAAAAJ&citation_for_view=HfiPj7oAAAAJ:eQOLeE2rZwMC"
target="_blank" rel="noopener noreferrer">Learning From Synthetic Labs: Language Models as Auction Participants</a>
<p class="pub__authors">Anand Shah, Kehang Zhu, Yanchen Jiang, Jeffrey G Wang, <strong>Arif Kerem Dayı</strong>, John J Horton, David C Parkes</p>
<p class="pub__venue">ArXiv Preprint <span>2025</span></p>
</li>
<li class="pub">
<a class="pub__title" href="https://scholar.google.com/citations?view_op=view_citation&hl=en&user=HfiPj7oAAAAJ&citation_for_view=HfiPj7oAAAAJ:2osOgNQ5qMEC"
target="_blank" rel="noopener noreferrer">Projected push-pull for distributed constrained optimization
over time-varying directed graphs</a>
<p class="pub__authors">Orhan Eren Akgun*, <strong>Arif Kerem Dayı*</strong>, Stephanie Gil, Angelia
Nedich</p>
<p class="pub__venue">American Control Conference (ACC) <span>2024</span></p>
</li>
<li class="pub">
<a class="pub__title" href="https://scholar.google.com/citations?view_op=view_citation&hl=en&user=HfiPj7oAAAAJ&citation_for_view=HfiPj7oAAAAJ:d1gkVwhDpl0C"
target="_blank" rel="noopener noreferrer">Learning trust over directed graphs in multiagent systems</a>
<p class="pub__authors">Orhan Eren Akgun, <strong>Arif Kerem Dayı</strong>, Stephanie Gil, Angelia
Nedich</p>
<p class="pub__venue">Learning for Dynamics and Control Conference (L4DC) <span>2023</span></p>
</li>
</ol>
</section>
<section id="experience" class="section">
<h2 class="section__title"><span class="section__no">03</span>Industry Experience</h2>
<ol class="timeline">
<li class="tl">
<span class="tl__date">Summer 2026</span>
<div class="tl__body">
<p class="tl__role">Research Intern</p>
<p class="tl__org">Microsoft Research - Cambridge, MA</p>
<!-- <p class="tl__desc"></p> -->
</div>
</li>
<li class="tl">
<span class="tl__date">Summer 2022</span>
<div class="tl__body">
<p class="tl__role">Data Science Intern</p>
<p class="tl__org">Trendyol Group - Istanbul, Turkey</p>
<!-- <p class="tl__desc">Worked on language modeling for unstructured text data analysis in e-commerce applications.</p> -->
</div>
</li>
</ol>
</section>
<section id="teaching" class="section">
<h2 class="section__title"><span class="section__no">04</span>Teaching</h2>
<p class="section__lede">During my time at Harvard, I had the privilege of
serving the CS and math community as a course assistant for:</p>
<ul class="courses">
<li>
<span class="course__code">CS 121</span>
<div>
<p class="course__name">Introduction to Theoretical Computer Science</p>
<p class="course__desc">Theory of computation — circuits, Turing machines,
computability, complexity, reductions, and randomized computation.
Taught by Boaz Barak.</p>
</div>
</li>
<li>
<span class="course__code">Math 25b</span>
<div>
<p class="course__name">Theoretical Linear Algebra & Real Analysis II</p>
<p class="course__desc">Rigorous single- and multivariable real analysis,
metric space topology, and basic Fourier analysis. Taught by Wes Cain.</p>
</div>
</li>
<li>
<span class="course__code">Math 25a</span>
<div>
<p class="course__name">Theoretical Linear Algebra & Real Analysis I</p>
<p class="course__desc">A rigorous introduction to linear algebra based on
the wonderful
<a href="https://linear.axler.net" target="_blank" rel="noopener noreferrer">Linear Algebra Done
Right</a>. Taught by Wes Cain.</p>
</div>
</li>
<li>
<span class="course__code">ES 150</span>
<div>
<p class="course__name">Probability with Engineering Applications</p>
<p class="course__desc">Fundamental probability concepts and their
applications to engineering problems. Taught by Yue M. Lu.</p>
</div>
</li>
</ul>
</section>
<section id="writing" class="section">
<h2 class="section__title"><span class="section__no">05</span>Writing</h2>
<p class="section__lede">Some expository notes and essays</p>
<ul class="writing">
<li class="note">
<a class="note__title" href="files/18_676_final_project.pdf" target="_blank" rel="noopener noreferrer">Stochastic
Localization and Sampling via Time Reversal</a>
<span class="note__date">2026</span>
<p class="note__desc">An expository class project (MIT 18.676) connecting sampling,
stochastic localization, and filtering through time-reversals of SDEs,
with an eye toward diffusion-based generative modeling.</p>
</li>
<li class="note">
<a class="note__title" href="https://dash.harvard.edu/entities/publication/ddd3e67b-c1f0-4632-971f-fb1cb28201a2" target="_blank" rel="noopener noreferrer">Learning in Neural Networks: Lazy training, Feature Learning, and Fine-Tuning</a>
<span class="note__date">2025</span>
<p class="note__desc">My undergraduate thesis including exposition on learning in neural networks and original research on learning dynamics of low-rank fine tuning.</p>
</li>
<li class="note">
<a class="note__title" href="files/cs224_final_project.pdf" target="_blank" rel="noopener noreferrer">Leap
Complexity: How SGD Exploits Hierarchical Structure to Learn
Efficiently</a>
<span class="note__date">Dec 2023</span>
<p class="note__desc">An expository class project (Harvard CS 224) on how stochastic gradient descent
can go beyond the kernel regime to perform hierarchical feature learning.</p>
</li>
</ul>
</section>
</main>
</div>
<footer class="footer">
<div class="footer__inner">
<span>© <span id="year">2026</span> Arif Kerem Dayı</span>
</div>
</footer>
<script data-cfasync="false" src="/cdn-cgi/scripts/5c5dd728/cloudflare-static/email-decode.min.js"></script><script>
// Active-section highlighting for the top nav.
const navLinks = [...document.querySelectorAll('[data-nav]')];
const byId = id => navLinks.find(a => a.getAttribute('href') === '#' + id);
const observer = new IntersectionObserver((entries) => {
entries.forEach(e => {
const link = byId(e.target.id);
if (link && e.isIntersecting) {
navLinks.forEach(a => a.classList.remove('is-active'));
link.classList.add('is-active');
}
});
}, { rootMargin: '-45% 0px -50% 0px' });
document.querySelectorAll('main .section').forEach(s => observer.observe(s));
document.getElementById('year').textContent = new Date().getFullYear();
</script>
<script type="module" src="https://static.cloudflareinsights.com/beacon.min.js/v4513226cdae34746b4dedf0b4dfa099e1781791509496" integrity="sha512-ZE9pZaUXND66v380QUtch/5sE9tPFh2zg45pR2PB0CVkCtOREv2AJKkSidISWkysEuQ0EH8faUU5du78bx87UQ==" data-cf-beacon='{"version":"2024.11.0","token":"53f1c3deadda404aad6d5aa4ef3c6b11","r":1}' crossorigin="anonymous"></script>
</body>
</html>
Cloudflare