A single neural network layer applies the weight matrix
to its input, with no bias and no activation function. Two training examples arrive as 2-dimensional embeddings:
| example | embedding |
|---|---|
| "cat" | |
| "dog" |
After the layer, the Euclidean distance between the two representations has been multiplied by what factor?
Once you get it right, the reasoning behind the answer shows up here.
Select all that apply.