# Mini Batch Gradient Descent Backpropapagation

I am a beginner to machine learning. I have derived the equations for backpropagation, and for the weight update for hidden layers, the update rule uses the output vector of the layer to multiply with the error from the next layer.

My question is, in mini-batch backpropagation (with, let's say, 200 samples), which sample's output do we use? For each sample, every layer has a different output, so when we run the backpropagation algorithm, which sample's output do we use?