CINXE.COM
Hannaneh Hajishirzi - University of Washington
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <title>Hannaneh Hajishirzi - University of Washington</title> <meta name="Keywords" content="" /> <meta name="Description" content="" /> <link href="default.css" rel="stylesheet" type="text/css" /> <script type="text/javascript"> var _gaq = _gaq || []; _gaq.push(['_setAccount', 'UA-21070967-1']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); </script> </head> <body> <div id="header"> <!--div id="colOne"><img src="images/UILogoCL2c.gif"/></div--> <div id="colTwo"><h1>Hannaneh Hajishirzi</h1></div> </div> <div id="content"> <div id="menu"> <ul> <li><a href="index.html" accesskey="1">Home</a></li> <li><a href="research.html" accesskey="2">Publications</a></li> <li><a href="teaching.html" accesskey="3">Teaching</a></li> <li><a href="https://h2lab.cs.washington.edu/members.html" accesskey="3">Students</a></li> <li><a href="http://h2lab.cs.washington.edu" accesskey="2">H2Lab</a></li> <!--li><a href="projects.html" accesskey="4">Projects</a></li--> <!--li><a href="bio.html" accesskey="5">Bio/CV</a></li--> </ul> </div> <div id="colThree"> <img src="hanna-webpage.jpg"/ width="40%" height="40%" float="left"> </div> <div id="colThree"> <font size ="2.5", color="black"> <p style="font-family:verdana"> Torode Family Associate Professor <br/> <a href="https://www.cs.washington.edu/"><font size="2.5" color="DarkBlue">Paul G. Allen School of Computer Science and Engineering</a><br/></font> <font size="2.5", color="black"> Adjunct at: UW Electrical and Computer Engineering, Linguistics </font> <br/> <a href="https://www.washington.edu/"><font size="2.5" color="DarkBlue">University of Washington</a><br/> <br /></font> <br><font size="2.5", color="black"> Senior Director, AllenNLP </font> <br/> <a href="https://allenai.org/"><font size ="2.5" color="DarkBlue"> Allen Institute for AI</a></font><br/> <br /></font> <font size="2.5"><strong>Email:</strong> hannaneh [at] cs [dot] washington [dot] edu<br/></font> <!--font size="2.5" color="navy"><a "font_color=black" href = "https://scholar.google.com/citations?user=LOV6_WIAAAAJ&hl=en">Google scholar, </a><a href="">, Twitter</a> <a href=" ">, Group Webpage (H2lab) </a> </p--> </div> </div--> <div id="colThree"> <font size="4"><h3> About </h3> </font> <font size="2.5"><p> Hanna Hajishirzi is the Torode Family Associate Professor in the Allen School of Computer Science and Engineering at the University of Washington and a Senior Director of NLP at AI2. She received her Ph.D in Computer Science from University of Illinois at Urbana-Champaign, and spent a year as Postdoctoral associate at Disney Research and CMU. Her current research delves into various areas within NLP and AI, with a particular focus on understanding and pushing the boundaries of large language models. She has published more than 140 scientific articles in top-tier journals and conferences in ML, AI, NLP, and Computer Vision. She is a recipient of 2020 Alfred Sloan Fellowship, 2021 NSF CAREER award, 2019 Intel rising star award, 2018 Allen Distinguished Investigator award, 2023 Academic Achievement UIUC Alumni award, 2024 innovator of the year award finalist by GeekWire, and several research faculty awards from industry. The work from her lab has been nominated or received best paper awards at conferences and have been featured in a variety of magazines and newspapers including New York Times, Forbes, NPR, MIT Technology Review, Geekwire, Wired Magazine, and more. </font></p> <font size="2.5"><p style="font-family:verdana"> <b>Recent awards: </b></p> <ul> <li> UIUC Academic Achievement Alumni Award </li> <li> NSF CAREER award </li> <li> Sloan Fellowship</li> <li> Intel Rising Star Faculty Award</li> <li> Allen Distinguished Investigator Award </li> <li> Research faculty awards: Amazon, Facebook, Google, Samsung GRO, Bloomberg </li> </ul> <font size="2.5"><p> My lab (<a href="http://h2lab.cs.washington.edu">H2lab</a>) mainly publishes at NLP (ACL, NAACL, EMNLP), AI and ML conferences (AAAI, ICLR) across these areas: </p> <font size="4"><h3> Research </h3> </font> My research is mainly focused on NLP and language modeling. The goals of my research include (1) establishing the science of language modeling through the OLMo project, (2) advancing their scope to make them applicable and useful for human lives through our post-training efforts, and (3) introducing a new generation of LMs that address fundamental challenges inherent in the current models through our retrieval-based langauge modeling efforts. <!--Effectively unlimited quantities of ever-changing knowledge are available online in diverse styles (e.g., news v. science text) and formats (knowledge bases, web pages, and textual documents). My research addresses the challenge of enabling rich neural symbolic comprehension and reasoning given this diversity: how can we build AI systems that comprehend and combine evidence from various and evolving sources of textual and multi-modal knowledge to make complex inferences and draw logical conclusions? We build algorithms that balance three competing desiderata:<b> interpretable, robust with high performance</b>, and <b> efficient and scalable</b> in the following categories: </p> <ul> <li><b>General-purpose NLP. </b> Building NLP models that go beyond solving individual tasks and can learn new tasks from their descriptions or a few examples. </li><br> <li><b>Reasoning and question answering. </b> Building benchmarks and algorithms that offer rich natural language comprehension using open domain, multi-lingual, multi-hop, and interpretable reasoning; developing some of the first deep neural models for general reading comprehension (BiDAF), open domain QA, cross-lingual QA, multi-hop reasoning, and symbolic methods to solve math and geometry word problems.</li> <br> <li><b>Knowledge acquisition from multi-modal data. </b> Devising general high-performance algorithms to extract knowledge from textual and visual data; devloping some of the first work in extracting knowledge from scientific text.</li> <br> <li><b>Representation learning. </b> Integrating capabilities of symbolic representations into neural models to represent knowledge acquired from diverse structured and un-structured resources and forming knowledge-rich dense vectors to encode them; designing neural architectures that efficiently encode textual and visual data.</li> </ul>--> </div> <div id="colThree"> <font size="3.5"><h3>Contact</h3></font> <font size="2.5"> <p style="font-family:verdana"> <ul> <strong>Office:</strong> Paul Allen Center 654 <br /> <strong>Phone:</strong> (206) 221-3921 <br /> <strong>Email:</strong> hannaneh [at] cs [dot] washington [dot] edu <br /> </p></font> </ul> </div> <div style="clear: both;"> </div> </div> <div id="footer"> <p>Last modified © 2020 Hannaneh Hajishirzi</p> </div> </body> </html>