add Ntampaka hands-on

Lecture_Day3_Ntampaka/README.md

@@ -1 +1,22 @@
 # Day 3: symmetries and specialized architectures
+
+## For Wednesday’s lesson:
+
+### Activity 1:  [Convolutional Neural Networks](https://archive.stsci.edu/hello-universe/deepmerge)
+Before class, please:
+1. follow the [notebook link](https://spacetelescope.github.io/hellouniverse/notebooks/hello-universe/Classifying_JWST-HST_galaxy_mergers_with_CNNs/Classifying_JWST-HST_galaxy_mergers_with_CNNs.html)
+and download the notebook
+2. follow the [data link](https://archive.stsci.edu/hlsp/deepmerge) and download the 2-filter pristine fits file at the bottom of the page
+3. run the imports cell and ensure that it runs. If it does not, install any necessary software packages.
+
+Note: this activity is also available on the [TIKE science platform](https://timeseries.science.stsci.edu/hub/spawn) (select the default environment to spin up an AWS server)
+
+### Activity 2:  [Interpreting CNNs](https://archive.stsci.edu/hello-universe/interpretability)
+Before class, please:
+1. follow the [notebook link](https://spacetelescope.github.io/hellouniverse/notebooks/hello-universe/Interpreting_CNNs/Interpreting_CNNs.html)
+and download the notebook
+2. this notebook uses the same data as above - no additional data is needed.
+3. run the imports cell and ensure that it runs. If it does not, install any necessary software packages.
+
+### Activity 3:  [Graph Neural Networks](https://colab.research.google.com/github/jwuphysics/LSSTC-DSFP-Session-19/blob/main/day4/Graph%20Neural%20Networks.ipynb)
+No work is needed before class.