Suggestion: clarify “next word” vs “next token” in causal language modeling explanation

I noticed a small wording detail in the LLM Course, Section 1 “Transformer Models”, Chapter 4 “How do Transformers work?”.

In the explanation of causal language modeling, the text says:

> “predicting the next word in a sentence having read the n previous words”

As an English learner and beginner in LLMs, I initially understood this literally as “word prediction”. Later I realized that modern Transformer language models actually operate on tokens rather than true linguistic words.

Maybe a slightly more technically precise wording could be something like:

> “predicting the next token given the previous n tokens”

or perhaps adding a short note mentioning that “word” is being used as a beginner-friendly simplification.

I think the current explanation is approachable for newcomers, but this small clarification might help learners better connect language modeling concepts with tokenization later in the course.

Search the course source for the exact sentence in Section 1, “Transformer Models,” Chapter 4, “How do Transformers work?”. Read the surrounding causal language modeling explanation, then revise the wording or add the suggested clarification about tokens while preserving its beginner-friendly tone. Check the rendered chapter to confirm the explanation is clear and accurate.

Written by the indexing model from the issue text.