Fixing typos

public/index.html

@@ -228,7 +228,7 @@ <h2>Aggregating Multiple Images</h2>
 
   <figure id="layer-annotation-5" style="grid-column: text/page;"></figure>
 
-  <p>Not let’s jump to the other side of the atlas, where we can see many variations of text detectors. These will be useful when identifying classes such as "menu", "web site" or "book jacket".
+  <p>Now let’s jump to the other side of the atlas, where we can see many variations of text detectors. These will be useful when identifying classes such as "menu", "web site" or "book jacket".
 
   <figure id="layer-annotation-6" style="grid-column: text/page;"></figure>
 
@@ -417,7 +417,7 @@ <h4>Mixed4e</h4>
 
   <h2>Focusing on a Single Classification</h2>
 
-  <p>Looking at an atlas of all activations can be a little overwhelming, especially when you're trying to understand on how the network goes about ranking one particular class. For instance, let's investigate how the network classifies a "fireboat".
+  <p>Looking at an atlas of all activations can be a little overwhelming, especially when you're trying to understand how the network goes about ranking one particular class. For instance, let's investigate how the network classifies a "fireboat".
 
   <figure style="display: grid; grid-template-columns: 3fr 1fr; grid-gap: 20px;">
     <div data-lazy-image-aspect-ratio="1.335" data-lazy-image="fireboat-01.jpg" alt="fireboat"></div>
@@ -490,7 +490,7 @@ <h2>Further Isolating Classes</h2>
 
   <figure id="snorkel-scuba-diver" style="grid-column: page"></figure>
 
-  <p>We can do something similar with two other closely related classes, "frying pan" and "wok" — two classes where I would  be hard-pressed to identify any differences at first blush.
+  <p>We can do something similar with two other closely related classes, "frying pan" and "wok" — two classes where we would  be hard-pressed to identify any differences at first blush.
 
   <figure id="class-gradient-frying-pan" style="grid-column: text"></figure>
 
@@ -527,7 +527,7 @@ <h3>Surfacing Inner Properties of Models</h3>
 
   <h3>New interfaces</h3>
 
-  <p>Machine learning models are usually deployed as black boxes that automate a specific task, executing it on their own. But there’s a growing sense that there might be an alternate way for us to relate to them: that instead of increasing automating, they could augment. One vision of this augmentation that we find particularly compelling is the idea that the internal representations neural networks learn can be repurposed as tools <d-cite key="carter2017using,olah2015visualizing,johnson2018machine"></d-cite>. Already, we’ve seen exciting demonstrations of this in images 
+  <p>Machine learning models are usually deployed as black boxes that automate a specific task, executing it on their own. But there’s a growing sense that there might be an alternate way for us to relate to them: that instead of increasingly automating task, they could be used to augment a user. One vision of this augmentation that we find particularly compelling is the idea that the internal representations neural networks learn can be repurposed as tools <d-cite key="carter2017using,olah2015visualizing,johnson2018machine"></d-cite>. Already, we’ve seen exciting demonstrations of this in images 
   <d-cite key="isola2017image,loh2017toposketch,zhu2016generative"></d-cite>
   and music 
   <d-cite key="mccurry2018latent,roberts2018hierarchical"></d-cite>