Large Language Models (LLMs) working principles UPSC Prelims PYQ 2026
Which of the following statements with regard to Large Language Models (LLMs) used in machine learning is/are correct?
- LLMs assign probabilities to the next possible words and then pick the one with the highest probability.
- LLMs process data through mathematical optimisation to minimise prediction errors.
- LLMs produce unbiased outputs.
Select the answer using the code given below:
- 1 only
- 1 and 2 only
- 2 and 3 only
- 1, 2 and 3
Explanation
Statement 1 is correct
- Large Language Models (LLMs) are advanced generative AI models trained on massive text datasets to understand, generate and manipulate human language with high accuracy. They are primarily built using deep learning techniques and transformer architectures. LLMs predict text by calculating probabilities for possible next tokens/words, and selecting one based on probability (sometimes with controlled randomness).
Statement 2 is correct
- LLMs are trained using optimisation algorithms (like gradient descent) to reduce prediction error during training.
Statement 3 is incorrect
- LLMs can reflect biases present in training data, societal patterns, internet content, and model design choices. Hence, outputs may contain bias despite mitigation efforts.
PMF Concept Hack
|
Answer: (b) 1 and 2 only; Difficulty Level: Easy
