Which option correctly explains Gradient descent in a neural-network workflow? Use the application-oriented interpretation.

Correct Answer: D

Gradient descent updates parameters in the direction that reduces loss

The correct answer is gradient descent updates parameters in the direction that reduces loss. This matches the Deep Learning course topic: Gradient descent.