diff --git a/lab01/README.md b/lab01/README.md
index d854cfe..4176394 100644
--- a/lab01/README.md
+++ b/lab01/README.md
@@ -121,12 +121,6 @@ We use a pre-trained AI model called **BERT (Bidirectional Encoder Representatio
 
 ### How BERT Works
 
-<img src="images/self-attention.png" width="60%">
-
-[![Watch Another YouTube Video](http://img.youtube.com/vi/EOmd5sUUA_A/0.jpg)](https://www.youtube.com/watch?v=EOmd5sUUA_A)
-
-[![Watch my YouTube Video](http://img.youtube.com/vi/xI0HHN5XKDo/0.jpg)](https://www.youtube.com/watch?v=xI0HHN5XKDo)
-
 BERT reads the entire sentence at once and uses a mechanism called **self-attention** to focus on important words. For example:
 
 ```
@@ -134,6 +128,12 @@ Sentence: "The bank was robbed."
 BERT understands that "bank" refers to a financial institution because of the word "robbed."
 ```
 
+<img src="images/self-attention.png" width="60%">
+
+[![Watch Another YouTube Video](http://img.youtube.com/vi/EOmd5sUUA_A/0.jpg)](https://www.youtube.com/watch?v=EOmd5sUUA_A)
+
+[![Watch my YouTube Video](http://img.youtube.com/vi/xI0HHN5XKDo/0.jpg)](https://www.youtube.com/watch?v=xI0HHN5XKDo)
+
 ### Why Use BERT?
 
 - **Pre-trained**: BERT has already learned from millions of sentences, so it understands language well.