neuron that learns to be an and gate try is the forward pass, you need to press it for it to change its output if you want to try to train it to be other logic gates then change the target in the skeletons script. ive had it work with and, nand, or, and nor you can turn on turbo mode by pressing the green flag while holding shift, it is recommended to do this with a low learning rate FORWARD PASS a neuron has inputs, each of those inputs has weights. it multiplies each of the inputs by its respective weight and sums them all together before adding a bias. then it passes it through an activation function, which is a sigmoid here. LOSS CALCULATION theres a target output, in the case of an and gate its just input 1 * input 2 the error or loss is calculated as (target - output)^2 BACKWARD PASS / backpropagation / gradient descent find the partial derivative of the loss in relation to every weight and bias, subtract that multiplied by the learning rate from the weight or bias
