manaully remake neural network from example

Hi,

i am trying to manually calculate the output values for the neural network which is made in this example: https://nl.mathworks.com/help/comm/ug/training-and-testing-a-neural-network-for-llr-estimation.html. (the Network for the 16-QAM).

it has 2 inputs, 1 hidden layer with 8 neurons and an output layer with 4 neurons.

I try to calculate the output values manually. But i get different results than the calculated values from the script.

My script now:

    w1 = net.IW{1} %the input-to-hidden layer weights
    w2 = net.LW{2} %the hidden-to-output layer weights
    b1 = net.b{1} %the input-to-hidden layer bias
    b2 = net.b{2} %the hidden-to-output layer bias

    % Layer 1
    a1 = tansig(repmat(b1,1,size(x,2)) + w1*x); 
    % output
    n=repmat(b2,1,size(x,2)) + w2*a1;
    output= purelin(n);

What am i doing wrong or what do i forget in my calculation to get the correct values?

thanks in advance!

[soepblik]

someone that could help?

it has 1 hidden layer with relu activation

and the output layer has a pure linear activation.