From 1d38628abf47e055935c603fae80728e516c8b8b Mon Sep 17 00:00:00 2001 From: Artem Oboturov Date: Tue, 5 May 2026 13:30:01 +0200 Subject: [PATCH 1/2] Replace retired cdn.mathjax.org with jsdelivr CDN The cdn.mathjax.org CDN was retired in 2017 (see https://www.mathjax.org/cdn-shutting-down/), which broke all math rendering on the site. Switch to jsdelivr's MathJax v2 mirror, keeping v2 because the inline MathJax.Hub.Config is v2-API. --- _includes/head.html | 2 +- docs/autoregressive/index.html | 2 +- docs/flow/index.html | 2 +- docs/gan/index.html | 2 +- docs/index.html | 2 +- docs/introduction/index.html | 2 +- docs/vae/index.html | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) diff --git a/_includes/head.html b/_includes/head.html index 602e239..9d573ab 100644 --- a/_includes/head.html +++ b/_includes/head.html @@ -16,7 +16,7 @@ {% endif %} {% if site.data.options.mathjax %} - + {% endif %} + + + + + + + + {% endif %} - - {{ content }} diff --git a/docs/autoregressive/index.html b/docs/autoregressive/index.html index c7ef31e..388743b 100644 --- a/docs/autoregressive/index.html +++ b/docs/autoregressive/index.html @@ -18,7 +18,47 @@ - + + - -

We begin our study into generative modeling with autoregressive models. As before, we assume we are given access to a dataset of -dimensional datapoints . For simplicity, we assume the datapoints are binary, i.e., .

Representation

diff --git a/docs/flow/index.html b/docs/flow/index.html index a53ebc7..28cfe33 100644 --- a/docs/flow/index.html +++ b/docs/flow/index.html @@ -18,7 +18,47 @@ - + + - -

We continue our study over another type of likelihood based generative models. As before, we assume we are given access to a dataset of -dimensional datapoints . So far we have learned two types of likelihood based generative models:

    diff --git a/docs/gan/index.html b/docs/gan/index.html index f8e231e..1dee4a7 100644 --- a/docs/gan/index.html +++ b/docs/gan/index.html @@ -18,7 +18,47 @@ - + + - -

    We now move onto another family of generative models called generative adversarial networks (GANs). GANs are unique from all the other model families that we have seen so far, such as autoregressive models, VAEs, and normalizing flow models, because we do not train them using maximum likelihood.

    Likelihood-free learning

    diff --git a/docs/index.html b/docs/index.html index 59b781a..236e4b1 100644 --- a/docs/index.html +++ b/docs/index.html @@ -18,7 +18,47 @@ - + + - -

    These notes form a concise introductory course on deep generative models. They are based on Stanford CS236, taught by Stefano Ermon and Aditya Grover, and have been written by Aditya Grover, with the help of many students and course staff. The notes are still under construction! diff --git a/docs/introduction/index.html b/docs/introduction/index.html index 209fba1..c9b062f 100644 --- a/docs/introduction/index.html +++ b/docs/introduction/index.html @@ -18,7 +18,47 @@ - + + - -

    Intelligent agents are constantly generating, acquiring, and processing data. This data could be in the form of images that we capture on our phones, text messages we share with our friends, graphs that model diff --git a/docs/vae/index.html b/docs/vae/index.html index a8edcfd..8a3615d 100644 --- a/docs/vae/index.html +++ b/docs/vae/index.html @@ -18,7 +18,47 @@ - + + - -