var sms = '<div id="sms_bulletin">  <div id="sms_header">  <div class="do_not_print" id="arrow_left"><a href="?next=-39"><img src="//www2.math.ethz.ch/smsjs/images/arrow_left.png"/></a></div><div class="do_not_print" id="arrow_right"><a href="?next=-37"><img src="//www2.math.ethz.ch/smsjs/images/arrow_right.png"/></a></div>    <div class="sms_title">      <span style="float: left;">SMS</span>      <span>SWISS MATHEMATICAL SOCIETY</span>      <span style="float: right;">SMG</span>    </div>    <div class="sms_info">      <span style="float: left; width: 250px;">Bulletin no. 9</span>      <span style="float: right; text-align:right; width: 250px;">29.7 - 2.8.2024</span>      <span style="text-align: center; display: block; width: 250px; margin: auto; position: relative;">Summer Break 2024</span>    </div>  </div>  <div></div>  <div id="sms_content">    <div class="day">      <div class="day_title">        Monday, July 29      </div>      <div class="day_content">        <div class="day_content_item normal"><div class="event_location uni_zh"><a class="external" href="https://ethz.ch">ETH Z&uuml;rich</a></div><div class="event_type"><a class="external" href="https://stat.ethz.ch/events/research_seminar">Research Seminar in Statistics</a></div><br /><div class="event_speaker"> Chi  Jin  (Princeton University)</div><div class="event_title">Beyond Equilibrium Learning <span class="abstract-button do_not_print">abstract</span></div><div class="abstract-text" style="display: none"><strong>Abstract:</strong><br />Abstract: While classical game theory primarily focuses on finding equilibria, modern machine learning applications introduce a series of new challenges where standard equilibrium notions are no longer sufficient, and the development of new efficient algorithmic solutions is urgently needed. In this talk, we will demonstrate two such scenarios: (1) a natural goal in multiagent learning is to learn rationalizable behavior, which avoids iteratively dominated actions. Unfortunately, such rationalizability is not guaranteed by standard equilibria, especially when approximation errors are present. Our work presents the first line of efficient algorithms for learning rationalizable equilibria with sample complexities that are polynomial in all problem parameters, including the number of players; (2) In multiplayer symmetric constant-sum games like Mahjong or Poker, a natural baseline is to achieve an equal share of the total reward. We demonstrate that the self-play meta-algorithms used by existing state-of-the-art systems can fail to achieve this simple baseline in general symmetric games. We will then discuss the new principled solution concept required to achieve this goal.Bio: Chi Jin is an assistant professor at the Electrical and Computer Engineering department of Princeton University. He obtained his PhD degree in Computer Science at University of California, Berkeley, advised by Michael I. Jordan. His research mainly focuses on theoretical machine learning, with special emphasis on nonconvex optimization and Reinforcement Learning (RL). In nonconvex optimization, he provided the first proof showing that first-order algorithm (stochastic gradient descent) is capable of escaping saddle points efficiently. In RL, he provided the first efficient learning guarantees for Q-learning and least-squares value iteration algorithms when exploration is necessary. His works also lay the theoretical foundation for RL with function approximation, multi-agency and partial observability. He received NSF CAREER award and Sloan fellowship.</div><div class="event_time">10:00 &bull; ETH Zentrum, R&auml;mistrasse 101, Z&uuml;rich, Building HG, Room G 19.1</div>        </div>      </div>    </div>    <div class="day">      <div class="day_title">        Tuesday, July 30      </div>      <div class="day_content">        <div class="day_content_item no_event">        No events scheduled today!        </div>      </div>    </div>    <div class="day">      <div class="day_title">        Wednesday, July 31      </div>      <div class="day_content">        <div class="day_content_item no_event">        No events scheduled today!        </div>      </div>    </div>    <div class="day">      <div class="day_title">        Thursday, August 1      </div>      <div class="day_content">        <div class="day_content_item no_event">        No events scheduled today!        </div>      </div>    </div>    <div class="day">      <div class="day_title">        Friday, August 2      </div>      <div class="day_content">        <div class="day_content_item no_event">        No events scheduled today!        </div>      </div>    </div>  </div>  <div id="sms_footer">    address: Alessandra Naldi Windler, Departement Mathematik, HG G 51.4, ETH Zurich, Raemistrasse 101, CH-8092 Zurich<br />    email: <a href="mailto:bulletin@math.ch">bulletin@math.ch</a> | phone: +41 44 632 3684 | url: <a href="https://math.ch">https://math.ch</a>    <br /><br /><img src="//www2.math.ethz.ch/smsjs/sms_logo.gif" width="150" alt="SMS/SMG Logo" title="SMS/SMG Logo" />  </div></div>';