Check your understanding of this module before moving on.
?
0/3
Checkpoint quiz 3 questions
A quick check on this module's key ideas before you move on
Question 01
What mathematical rule is backpropagation fundamentally an application of?
A is correct. Backpropagation computes how the loss changes with respect to every weight by repeatedly applying the chain rule, propagating the gradient of the loss backward through each layer.
Question 02
Why does backpropagation compute gradients starting from the output layer and moving backward, rather than starting from the input?
A is correct. Each layer's gradient depends on the gradient of the layer after it, via the chain rule. That dependency only resolves cleanly if you start at the loss (the output) and work backward.
Question 03
What does it mean, in practice, when backpropagation produces a gradient of exactly zero for a particular weight?
B is correct. A zero gradient simply means the loss is locally flat with respect to that weight right now. It doesn't imply the model is done training, just that this particular weight won't move on this particular update.