Google Cloud Certified Generative AI Leader

Question 1 of 30

Fundamentals of generative AI

Which underlying mechanism primarily allows modern Large Language Models (LLMs) to generate coherent and contextually relevant text sequences?

They rely on supervised learning with rigidly structured tabular datasets to map specific inputs to exact, predefined outputs.

They predict the next most likely token in a sequence based on the patterns, contexts, and probabilities learned from vast amounts of training data.

They search a massive database of pre-written responses and retrieve the most relevant match based on keyword overlap.

They use reinforcement learning algorithms exclusively to logically deduce the factual answer to a user's prompt.