Guys I got a question. For example I have a layer_1 with 3 neurons. Then I pass some input there, and receieve a2 with 3 variables (0.9,0.7,0.6), but then I have a layer_2 with 10 neurons. How can I pass the variables to the layer_2. Like I am gonna have 30 variables in array or what? I didn’t get that clearly

Hello @karibay

Welcome to the community.

We have the weight matrices with dimensions/shape that will bridge layer 1 and layer 2. Check out this article and see if it answers your question,

This is so called “fully connected” layer, i.e, everything is connected like this.

Your output from the neuron 1 in the layer 1 goes to all neurons in the layer 2, but with different weights. And, those weights are trainable. So, depending to expected outputs, the network is trained to have appropriate weights among neurons.

And, the number of connections among two layers is 30 as you calculated, but all are vectorized and simplified as described in the link that shanup provided.

Hope this helps.

If you have a dataset of 100 rows and 10 columns and layer_1 has **n** neurons then you get

100 rows and n columns output