Quiz – LLM Fundamentals Explained

Question 1 of 5

1. Question

In which year was the Transformer architecture that powers LLMs introduced?

2017

Correct Incorrect Correct answer

1950

Correct Incorrect Correct answer

2005

Correct Incorrect Correct answer

2020

Correct Incorrect Correct answer

Question 2 of 5

2. Question

How do LLMs generate text?

Through token prediction, one piece at a time based on probability distributions

Correct Incorrect Correct answer

By copying complete sentences directly from their training set

Correct Incorrect Correct answer

By applying fixed grammar rules with no probabilities

Correct Incorrect Correct answer

By reasoning through formal logic before every word

Correct Incorrect Correct answer

Question 3 of 5

3. Question

What are tokens, according to the capsule?

Building blocks of language such as words, punctuation, or subword units

Correct Incorrect Correct answer

Physical memory chips inside the server

Correct Incorrect Correct answer

Security passwords used to access the model

Correct Incorrect Correct answer

Labels attached to supervised training examples

Correct Incorrect Correct answer

Question 4 of 5

4. Question

What is the role of attention mechanisms in the Transformer architecture?

They allow models to focus selectively on relevant context

Correct Incorrect Correct answer

They store the final answer for each possible question

Correct Incorrect Correct answer

They translate tokens back into physical text on screen

Correct Incorrect Correct answer

They guarantee that outputs are always factually accurate

Correct Incorrect Correct answer

Question 5 of 5

5. Question

What do scaling laws demonstrate about LLMs?

Larger models trained on more data generally perform better

Correct Incorrect Correct answer

Smaller models always outperform larger ones

Correct Incorrect Correct answer

Model size has no effect on performance

Correct Incorrect Correct answer

Adding more data reliably eliminates all hallucinations

Correct Incorrect Correct answer