{"id":11021,"date":"2018-04-23T06:07:49","date_gmt":"2018-04-23T10:07:49","guid":{"rendered":"http:\/\/whamit.mit.edu\/?p=11021"},"modified":"2018-04-21T12:13:11","modified_gmt":"2018-04-21T16:13:11","slug":"complang-423-hendrik-strobelt-ibm","status":"publish","type":"post","link":"https:\/\/ling-phil.mit.edu\/whamit\/2018\/04\/23\/complang-423-hendrik-strobelt-ibm\/","title":{"rendered":"CompLang 4\/23 - Hendrik Strobelt (IBM)"},"content":{"rendered":"<p><strong>Speaker:<\/strong>\u00a0<a href=\"http:\/\/hendrik.strobelt.com\/\">Hendrik Strobelt<\/a> (IBM)<br clear='none'\/>\n <strong>Title:\u00a0<\/strong> Visualization for Sequence Models for Debugging and Fun<br clear='none'\/>\n <strong>Date and time:<\/strong> Monday, April 23, 5-6pm<br clear='none'\/>\n <strong>Location:<\/strong> 46-5165<br clear='none'\/>\n <strong>Abstract:<\/strong><\/p>\n<blockquote>\n<p>Visual analysis is a great tool to explore deep learning models when there is no strong mathematical hypothesis yet<br clear='none'\/>\n available. I will present two visual tools where we used design study methodology to allow exploration of<br clear='none'\/>\n patterns in hidden state changes in RNNs\/LSTMs (LSTMVis) and exploration of Sequence2Sequence models (Seq2Seq-Vis).<br clear='none'\/>\n Both model types have shown superior performance for NLP like language modeling or language translation.<br clear='none'\/>\n Examples about both tasks will be shown on a variety of models.<\/p>\n<p>As beautiful distraction, we also utilize data science methods to investigate large data in a more artistic way.<br clear='none'\/>\n Formafluens is such a data experiment where we analyze a large collections of doodles made by humans in the Google<br clear='none'\/>\n Quickdraw tool.<\/p>\n<\/blockquote>\n","protected":false},"excerpt":{"rendered":"<p>Speaker:\u00a0Hendrik Strobelt (IBM) Title:\u00a0 Visualization for Sequence Models for Debugging and Fun Date and time: Monday, April 23, 5-6pm Location: 46-5165 Abstract: Visual analysis is a great tool to explore deep learning models when there is no strong mathematical hypothesis yet available. I will present two visual tools where we used design study methodology to [&hellip;]<\/p>\n","protected":false},"author":17,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[],"class_list":["post-11021","post","type-post","status-publish","format-standard","hentry","category-talks"],"_links":{"self":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/11021","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/users\/17"}],"replies":[{"embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/comments?post=11021"}],"version-history":[{"count":2,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/11021\/revisions"}],"predecessor-version":[{"id":11028,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/11021\/revisions\/11028"}],"wp:attachment":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/media?parent=11021"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/categories?post=11021"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/tags?post=11021"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}