uiz Space

May 2026 term · Deep Learning · BSCS3004

Deep Learning Quiz 2: 16 August 2026 (May 2026 term)

The IIT Madras BS Deep Learning (Deep Learning) Quiz 2 paper sat on 16 Aug 2026, in the May 2026 term: 19 questions for 50 marks in 120 minutes. Every question is below with its answer. Take it as a timed mock test to be marked, or read it through first.

Questions
19
Marks
50
Duration
120 min
MCQ
10
Numerical
7
MSQ
2

Updated

Official paper: Deep Learning 16 Aug 26 · No negative marking.

Question 1

+3 marksOne correct option
  1. A

    Bias will decrease significantly, while variance remains approximately the same.

  2. B

    Variance will tend to decrease, while bias remains approximately the same.

  3. C

    Both bias and variance will eventually become zero.

  4. D

    Increasing the amount of training data has no effect on either the bias or the variance.

Show answer

Correct answer

  • B

    Variance will tend to decrease, while bias remains approximately the same.

Question 2

+3 marksOne correct option
  1. A
  2. B
  3. C
  4. D

    At all iterations

Show answer

Correct answer

  • B

Question 3

+3 marksNumerical answer

Based on the above data, answer the given subquestions.

Show answer

Correct answer: 0.0035 (accepted within ±0.0003)

Question 4

+3 marksOne correct option

Based on the above data, answer the given subquestions.

  1. A

    It would increase from iteration 1 to iteration 3.

  2. B

    It would decrease from iteration 1 to iteration 3.

  3. C
  4. D

    It would decrease toward zero.

Show answer

Correct answer

  • C

Question 5

+3 marksOne correct option

Based on the above data, answer the given subquestions.

  1. A
  2. B
  3. C
  4. D
Show answer

Correct answer

  • B

Question 6

+3 marksOne correct option

Based on the above data, answer the given subquestions.

  1. A

    0

  2. B
  3. C
  4. D
Show answer

Correct answer

  • B

Question 7

+3 marksNumerical answer

What is the expected number of units retained (active) in a single forward pass during training?

Show answer

Correct answer: 7

Question 8

+3 marksOne correct option

How many distinct ''thinned'' sub-networks are theoretically possible for this layer?

  1. A
  2. B
  3. C
  4. D
Show answer

Correct answer

  • C

Question 9

+2 marksNumerical answer
Show answer

Correct answer: -1

Question 10

+2 marksNumerical answer
Show answer

Correct answer: -0.1

Question 11

+2 marksOne correct option

Which of the following is the primary motivation for using Leaky ReLU instead of standard ReLU in this setting?

  1. A

    To make the activation function computationally cheaper than ReLU.

  2. B
  3. C
  4. D
Show answer

Correct answer

  • C

Question 12

+2 marksOne correct option
  1. A
  2. B
  3. C
  4. D
Show answer

Correct answer

  • B

Question 13

+2 marksOne correct option
  1. A
  2. B
  3. C
  4. D
Show answer

Correct answer

  • B

Question 14

+2 marksOne or more correct options

Based on the above data, answer the given subquestions.

Which filter weights can receive a non-zero gradient during backpropagation? (Select all that apply.)

Select all that apply.

  1. A
  2. B
  3. C
  4. D
Show answer

Correct answers

  • C
  • D

Question 15

+3 marksNumerical answer

Based on the above data, answer the given subquestions.

Suppose

.

Show answer

Correct answer: 2

Question 16

+3 marksNumerical answer
Show answer

Correct answer: 3

Question 17

+2 marksOne correct option
  1. A
  2. B
  3. C
  4. D
  5. E
Show answer

Correct answer

  • A

Question 18

+2 marksNumerical answer
Show answer

Correct answer: 592

Question 19

+4 marksOne or more correct options

Which of the following statements are TRUE about the adaptive optimizers? Select all that apply.

Select all that apply.

  1. A
  2. B
  3. C
  4. D
  5. E

    (A figure from the original paper is missing from the source site.)

Show answer

Correct answers

  • B
  • E

    (A figure from the original paper is missing from the source site.)