{"version":"1.0","provider_name":"Microsoft Research","provider_url":"https:\/\/www.microsoft.com\/en-us\/research","author_name":"Xiaodong Liu","author_url":"https:\/\/www.microsoft.com\/en-us\/research\/people\/xiaodl\/","title":"Bridging Discrete and Backpropagation: Straight-Through and Beyond - Microsoft Research","type":"rich","width":600,"height":338,"html":"<blockquote class=\"wp-embedded-content\" data-secret=\"wlgAcZ7P79\"><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/publication\/bridging-discrete-and-backpropagation-straight-through-and-beyond\/\">Bridging Discrete and Backpropagation: Straight-Through and Beyond<\/a><\/blockquote><iframe sandbox=\"allow-scripts\" security=\"restricted\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/publication\/bridging-discrete-and-backpropagation-straight-through-and-beyond\/embed\/#?secret=wlgAcZ7P79\" width=\"600\" height=\"338\" title=\"&#8220;Bridging Discrete and Backpropagation: Straight-Through and Beyond&#8221; &#8212; Microsoft Research\" data-secret=\"wlgAcZ7P79\" frameborder=\"0\" marginwidth=\"0\" marginheight=\"0\" scrolling=\"no\" class=\"wp-embedded-content\"><\/iframe><script type=\"text\/javascript\">\n\/* <![CDATA[ *\/\n\/*! This file is auto-generated *\/\n!function(d,l){\"use strict\";l.querySelector&&d.addEventListener&&\"undefined\"!=typeof URL&&(d.wp=d.wp||{},d.wp.receiveEmbedMessage||(d.wp.receiveEmbedMessage=function(e){var t=e.data;if((t||t.secret||t.message||t.value)&&!\/[^a-zA-Z0-9]\/.test(t.secret)){for(var s,r,n,a=l.querySelectorAll('iframe[data-secret=\"'+t.secret+'\"]'),o=l.querySelectorAll('blockquote[data-secret=\"'+t.secret+'\"]'),c=new RegExp(\"^https?:$\",\"i\"),i=0;i<o.length;i++)o[i].style.display=\"none\";for(i=0;i<a.length;i++)s=a[i],e.source===s.contentWindow&&(s.removeAttribute(\"style\"),\"height\"===t.message?(1e3<(r=parseInt(t.value,10))?r=1e3:~~r<200&&(r=200),s.height=r):\"link\"===t.message&&(r=new URL(s.getAttribute(\"src\")),n=new URL(t.value),c.test(n.protocol))&&n.host===r.host&&l.activeElement===s&&(d.top.location.href=t.value))}},d.addEventListener(\"message\",d.wp.receiveEmbedMessage,!1),l.addEventListener(\"DOMContentLoaded\",function(){for(var e,t,s=l.querySelectorAll(\"iframe.wp-embedded-content\"),r=0;r<s.length;r++)(t=(e=s[r]).getAttribute(\"data-secret\"))||(t=Math.random().toString(36).substring(2,12),e.src+=\"#?secret=\"+t,e.setAttribute(\"data-secret\",t)),e.contentWindow.postMessage({message:\"ready\",secret:t},\"*\")},!1)))}(window,document);\n\/* ]]> *\/\n<\/script>\n","description":"Backpropagation, the cornerstone of deep learning, is limited to computing gradients solely for continuous variables. This limitation hinders various research on problems involving discrete latent variables. To address this issue, we propose a novel approach for approximating the gradient of parameters involved in generating discrete latent variables. First, we examine the widely used Straight-Through (ST) [&hellip;]"}