How chatbots work: quick check

Question 1

At its core, what does a large language model do?

Accepted Answer

Predicts the next token, over and over

Answer

Searches a database of correct answers

Answer

Copies answers from a human operator

Question 2

What is a "hallucination" in AI?

Accepted Answer

When the AI confidently makes something up

Answer

When the AI shows images

Answer

When the AI refuses to answer

Question 3

Why might a chatbot not know about last week's news?

Accepted Answer

It has a knowledge cutoff from its training data

Answer

It is not allowed to read news

Answer

News is too long

Question 4

What is fine-tuning for?

Accepted Answer

Teaching the model to be helpful and safe in conversations

Answer

Making the model smaller

Answer

Adding more languages only