Standardize seealso admonitions with JAX template and intersphinx references

lectures/_config.yml

@@ -98,6 +98,9 @@ sphinx:
       advanced:
         - "https://python-advanced.quantecon.org"
         - null
+      jax:
+        - "https://jax.quantecon.org/"
+        - null
     mathjax3_config:
       tex:
         macros:

lectures/newton_method.md

@@ -29,8 +29,7 @@ kernelspec:
 ```
 
 ```{seealso}
-**GPU:** A version of this lecture which makes use of [jax](https://jax.readthedocs.io) to run the code
-on a `GPU` is [available here](https://jax.quantecon.org/newtons_method.html)
+A version of this lecture using [JAX](https://github.com/jax-ml/jax) is {doc}`available here <jax:newtons_method>`
 ```
 
 ## Overview
@@ -790,7 +789,7 @@ With the larger overhead, the speed is not better than the optimized `scipy` fun
 Our next step is to investigate a large market with 3,000 goods.
 
 A JAX version of this section using GPU accelerated linear algebra and
-automatic differentiation is available [here](https://jax.quantecon.org/newtons_method.html#application)
+automatic differentiation is {doc}`available here <jax:newtons_method>`
 
 The excess demand function is essentially the same, but now the matrix $A$ is $3000 \times 3000$ and the parameter vectors $b$ and $c$ are $3000 \times 1$.
 

lectures/wealth_dynamics.md

@@ -24,8 +24,7 @@ kernelspec:
 ```
 
 ```{seealso}
-A [version of this lecture](https://jax.quantecon.org/wealth_dynamics.html) using a `GPU`
-is [available here](https://jax.quantecon.org/wealth_dynamics.html)
+A version of this lecture using [JAX](https://github.com/jax-ml/jax) is {doc}`available here <jax:wealth_dynamics>`
 ```
 
 In addition to what's in Anaconda, this lecture will need the following libraries: