Which of the following statements with regard to Large Language Models (LLMs) used in machine learning is/are correct? 1. LLMs as…

Question

Which of the following statements with regard to Large Language Models (LLMs) used in machine learning is/are correct?

1. LLMs assign probabilities to the next possible words and then pick the one with the highest probability.

2. LLMs process data through mathematical optimization to minimise prediction errors.

3. LLMs produce unbiased outputs.

1. A. 1 only
2. B. 1 and 2 only (Correct answer)
3. C. 2 and 3 only
4. D. 1, 2 and 3

Correct Answer

Option B — B

Detailed Solution & Explanation

The correct answer is Option B.

Key Points

• Statement 1 is correct: a Large Language Model predicts text by assigning probabilities to the next possible token/word and selecting a high-probability continuation (with some controlled randomness in practice).
• Statement 2 is correct: LLMs are trained by mathematical optimisation — gradient descent that minimises a prediction-error (loss) function across vast text data.
• Statement 3 is incorrect, and is the key: LLMs do not produce unbiased outputs. Because they learn from human-generated data, they can reproduce and even amplify the biases present in that data.
• LLMs are built on the transformer architecture and 'learn' statistical patterns of language; they have no understanding or intent, and their outputs reflect their training data — including its biases.

Exam Tip

• any claim that an AI system is inherently 'unbiased', 'objective' or 'always correct' is almost always the planted false statement.