GATE DA 2025 — Question 42

Supervised Learningmulti-layer perceptron, feed-forward neural networkMCQ2 marks · −2/3 if wrongMedium

Consider the neural network shown in the figure with inputs: \(u, v\) weights: \(a, b, c, d, e, f\) output: \(y\) \(R\) denotes the ReLU function, \(R(x) = \max(0, x)\).
Question image
Given \(u = 2, v = 3\), \(a = 1, b = 1, c = 1, d = -1, e = 4, f = -1\), which one of the following is correct?
  1. A
    \(\dfrac{\partial y}{\partial a} = 8, \dfrac{\partial y}{\partial f} = 0\)
  2. B
    \(\dfrac{\partial y}{\partial a} = 1, \dfrac{\partial y}{\partial f} = 0\)
  3. C
    \(\dfrac{\partial y}{\partial a} = 1, \dfrac{\partial y}{\partial f} = -1\)
  4. D
    \(\dfrac{\partial y}{\partial a} = 2, \dfrac{\partial y}{\partial f} = -1\)
Practise the full 2025 paper
Challenge a friend: WhatsApp Telegram
From RENYXERAThis week's 10-question challengeSame 10 past-GATE questions for everyone this week — see where you stand.
Practise more on multi-layer perceptron, feed-forward neural networkTimed, in the real exam interface · free to start

A GATE question every morning

One real past-paper question every morning for your branch, answer hidden until you tap. Free on Telegram.

Join
Stuck? Ask in the student group