Quiz Space

Large Language Models · End Term · 31 Aug 2025 · May 2025 term · Set QDB1

Question 5: Which of the following are valid design choices when buil…

Question 5

+3 marksOne or more correct options

Which of the following are valid design choices when building a transformer-based language model?

Select all that apply.

  1. A

    Number of transformer layers

  2. B

    Learning rate schedule

  3. C

    Font style of the source code

  4. D

    Vocabulary size

  5. E

    Type of normalization (e.g., LayerNorm)

Show answer

Correct answers

  • A

    Number of transformer layers

  • B

    Learning rate schedule

  • D

    Vocabulary size

  • E

    Type of normalization (e.g., LayerNorm)

See the whole paperTake it as a mock test

Question 5 of 17 in the IIT Madras BS Large Language Models (LLM) End Term paper sat on 31 Aug 2025, in the May 2025 term (IIT M DEGREE AN EXAM QDB3 31 Aug 2025). It carries 3 marks.

More questions from this paper

  1. Q1Given the input string: moonlight And the following vocabulary of subword tokens with their corresponding log-probabili…
  2. Q2Figure question
  3. Q3Which of the following statements about subword tokenizers (e.g., Byte Pair Encoding, SentencePiece) is true?
  4. Q4A GPT-based model is fine-tuned for story generation. During evaluation, you test different decoding strategies on the …
  5. Q6Which of the following statements about Top-k and Nucleus (Top-p) sampling are correct?
  6. Q7Which of the following preprocessing steps are commonly required when preparing data for transformer models like BERT?
  7. Q8Suppose a language model outputs the following logits for a vocabulary of size 4: | Token | Logit | |---|---| | cat | 1…
  8. Q9For the input “you enjoy tea often”, compute the final representation of the word “tea” after the attention layer (i.e.…
  9. Q10Suppose the input sentence is “tea you enjoy often”. Using the same matrices and processing method, what is the attenti…
  10. Q11Figure question
  11. Q12Figure question
  12. Q13What percentage of entries in the attention matrix are zero (i.e., the sparsity)?
  13. Q14Consider the short sentence: “large language models are cool” A transformer processes this sequence using the naive rel…
  14. Q15Consider the short sentence: “large language models are cool” A transformer processes this sequence using the naive rel…
  15. Q16Consider the short sentence: “large language models are cool” A transformer processes this sequence using the naive rel…
  16. Q17Consider the short sentence: “large language models are cool” A transformer processes this sequence using the naive rel…