<div dir="ltr"><h3 class="gmail-pfblock-header" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 60px;font-size:13px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;text-align:center"><br class="gmail-Apple-interchange-newline">THE COMPUTATIONAL SOCIAL SCIENCE WORKSHOP PRESENTS</h3><h1 class="gmail-pfblock-header3" style="box-sizing:border-box;margin:0px 0px 10px;font-size:40px;font-family:Raleway,sans-serif;line-height:1.2;font-stretch:normal;text-transform:uppercase;letter-spacing:5px;padding:0em;text-align:center">NICOLAS MASSE</h1><h3 class="gmail-pfblock-header3" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 10px;font-size:13px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;padding:0em;text-align:center">POSTDOCTORAL SCHOLAR AND STAFF SCIENTIST, DEPARTMENT OF NEUROBIOLOGY</h3><h3 class="gmail-pfblock-header3" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 10px;font-size:13px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;padding:0em;text-align:center">THE UNIVERSITY OF CHICAGO</h3><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"><br style="box-sizing:border-box"></p><p class="gmail-pfblock-header3" style="box-sizing:border-box;margin:0px 0px 10px;padding:0em;text-align:center;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px">The<span class="gmail-Apple-converted-space"> </span><a href="https://macss.uchicago.edu/content/computation-workshop" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">Computational Social Science Workshop<span class="gmail-Apple-converted-space"> </span></a>at the University of Chicago cordially invites you to attend this week’s talk:</p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"></p><div class="gmail-pfblock-header3" style="box-sizing:border-box;padding:0em;text-align:center;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><h2 class="gmail-pfblock-header" style="box-sizing:border-box;line-height:1.2;color:rgb(34,34,34);margin:0px 0px 60px;font-size:30px;font-stretch:normal;text-transform:uppercase;letter-spacing:5px"><a href="https://github.com/uchicago-computation-workshop/nicolas_masse/blob/master/CatastrophicForgetting.pdf" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">TWO STORIES AT THE INTERSECTION BETWEEN NEUROSCIENCE AND ARTIFICIAL INTELLIGENCE<span class="gmail-Apple-converted-space"> </span></a></h2><p style="box-sizing:border-box;margin:0px 0px 10px"><br style="box-sizing:border-box"></p></div><p class="gmail-footertext2" style="box-sizing:border-box;margin:0px auto 10px;text-align:justify;font-size:14px;color:rgb(102,102,102);font-family:Raleway,sans-serif"></p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><span style="box-sizing:border-box;font-weight:700">Summary:</span><span class="gmail-Apple-converted-space"> </span>While the fields of neuroscience and (especially) artificial intelligence have rapidly progressed over the last decade, what is perhaps less appreciated is how intertwined these two fields have become. This presentation will discuss two studies that highlight the synergy between these fields of research. In the first part, we will discuss how recurrent neural network models can be used to generate novel hypotheses regarding the neural substrates of working memory in the brain. These network models suggest that the maintenance and manipulation of information in working memory engage different neural substrates, potentially addressing a long-standing debate in the field. In the second part (paper attached), we will discuss how insights from neuroscience can be used to create novel algorithms that allow neural networks to learn large numbers of tasks with little forgetting. We show that mimicking the brain’s strategy of employing multiple, complementary methods can be a highly efficient and effective approach to alleviate forgetting in neural networks.</p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"></p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"></p><h4 class="gmail-pfblock-header3" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 10px;font-size:12px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;padding:0em;text-align:center">THURSDAY, 1/17/2019</h4><h4 class="gmail-pfblock-header3" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 10px;font-size:12px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;padding:0em;text-align:center">11:00AM-12:20PM</h4><h4 class="gmail-pfblock-header3" style="box-sizing:border-box;font-family:Raleway,sans-serif;line-height:1.2;margin:0px 0px 10px;font-size:12px;font-stretch:normal;text-transform:uppercase;letter-spacing:1px;padding:0em;text-align:center">KENT 120</h4><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"></p><p class="gmail-pfblock-header3" style="box-sizing:border-box;margin:0px 0px 10px;padding:0em;text-align:center;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px">A light lunch will be provided by Jimmy John’s.</p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"><br style="box-sizing:border-box"></p><p class="gmail-footertext2" style="box-sizing:border-box;margin:0px auto 10px;text-align:justify;font-size:14px;color:rgb(102,102,102);font-family:Raleway,sans-serif"><span style="box-sizing:border-box;font-weight:700">Nicolas Masse</span><span class="gmail-Apple-converted-space"> </span>is a staff scientist in the Department of Neurobiology at the University of Chicago. His research focuses on developing neuroscience-inspired artificial intelligence, and is primarily interested in creating neural networks that can seamlessly switch between different contexts and environments. Before arriving in Chicago, Nicolas worked on developing brain-computer interface systems for people with tetraplegia at Brown University, researched the olfactory system at Cambridge, and studied the neural mechanisms underlying visual motion perception at McGill University. Prior to academia, Nicolas developed statistical models for a sports gambling company, and taught high school mathematics.</p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"></p><p style="box-sizing:border-box;margin:0px 0px 10px;color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><br style="box-sizing:border-box"><br style="box-sizing:border-box"></p><hr style="box-sizing:content-box;height:0px;margin-top:20px;margin-bottom:20px;border-right-width:0px;border-bottom-width:0px;border-left-width:0px;border-top-style:solid;border-top-color:rgb(238,238,238);color:rgb(102,102,102);font-family:Raleway,sans-serif;font-size:14px"><p class="gmail-footertext" style="box-sizing:border-box;margin:0px auto 10px;text-align:justify;font-size:10.8px;color:rgb(102,102,102);font-family:Raleway,sans-serif">The 2018-2019<span class="gmail-Apple-converted-space"> </span><a href="https://macss.uchicago.edu/content/computation-workshop" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">Computational Social Science Workshop<span class="gmail-Apple-converted-space"> </span></a>meets Thursdays from 11 a.m. to 12:20 p.m. in Kent 120. All interested faculty and graduate students are welcome.</p><p class="gmail-footertext" style="box-sizing:border-box;margin:0px auto 10px;text-align:justify;font-size:10.8px;color:rgb(102,102,102);font-family:Raleway,sans-serif">Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the<span class="gmail-Apple-converted-space"> </span><a href="https://github.com/uchicago-computation-workshop/nicolas_masse/issues" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">issues page<span class="gmail-Apple-converted-space"> </span></a>of the<span class="gmail-Apple-converted-space"> </span><a href="https://github.com/uchicago-computation-workshop/nicolas_masse" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">workshop’s public repository on GitHub.</a><span class="gmail-Apple-converted-space"> </span>Further instructions are documented in the Computational Social Science Workshop’s<span class="gmail-Apple-converted-space"> </span><a href="https://github.com/uchicago-computation-workshop/README" style="box-sizing:border-box;background-color:transparent;color:rgb(66,139,202);text-decoration-line:none">README on Github.</a></p><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div dir="ltr"><div style="font-size:12.8px;color:rgb(0,0,0);font-family:Helvetica"><div style="font-size:12.8px">Nora Nickels</div><div style="font-size:12.8px">Ph.D. Candidate, <span style="font-size:12.8px">Department of Comparative Human Development</span></div><div style="font-size:12.8px">Preceptor, Masters in Computational Social Science</div><div style="font-size:12.8px">Fellow, Institute for Mind and Biology</div><div style="font-size:12.8px">The University of Chicago</div></div></div></div></div></div></div></div>