Indian Institute of Technology Madras
BS Degree Programme · Examination & Study Paper
Deep Learning
RNNs, encoder-decoder & attention
Examination Instructions & Guidelines
- Solve on notebook/paper first before checking solutions.
- MCQ: Exactly 1 option is correct. Negative marking applies.
- MSQ: One or more options correct. Select all to score.
- NAT: Numerical response. No negative marking.
- Scientific calculator allowed for all calculations.
- Official answer key is sealed at the end of this sheet.
End Term · 31 Aug 2025 (AN)End Term
Read the passage below, then answer Q1–Q3

Based on the above data, answer the given subquestions.
Q1NAT2 Marks·Write the number
Your answer:- Q2NAT4 Marks·Write the number
Your answer: - Q3NAT2 Marks·Write the number
Your answer:
End Term · 13 Apr 2025 (AN)End Term
- Q4MSQ3 Marks·Choose all correct options
Consider the following tasks which can be solved using an encoder-decoder architecture. Which of these tasks has a decoder where an RNN is not necessary?
- AVideo classification: outputs a single class label
- BSentiment analysis: outputs a single class label
- CDocument summarization: outputs a sequence of text
- DMachine translation: outputs a sequence of text
- EVideo captioning: outputs a sequence of text
Read the passage below, then answer Q5–Q7

Based on the above data, answer the given subquestions.
Q5NAT2 Marks·Write the numberWith a total of 7 time steps (T = 7), what is the total count of parameters (including bias) within the network?
Your answer:- Q6MCQ2 Marks·Choose one option

- A

- B

- C

- D

- A
- Q7NAT2 Marks·Write the number
Your answer: