The gradient at each layer depends on the gradient propagated from the layer above it.
A very large gradient can cause excessively large weight updates, potentially making training unstable.
Small gradients always cause the loss to decrease during training.
Correct answers
The gradient at each layer depends on the gradient propagated from the layer above it.
A very large gradient can cause excessively large weight updates, potentially making training unstable.
Question 2 of 19 in the IIT Madras BS Deep Learning (Deep Learning) End Term paper sat on 13 Sept 2026, in the May 2026 term (Deep Learning 13 Sep 26 (Session 2)). It carries 3 marks.
