What is a Large Language Model and how does it generate text?

Q: What is a Large Language Model and how does it generate text?

**TL;DR.** An LLM is a **transformer** trained on massive text to **predict the next token**. Generation is **autoregressive**: it predicts one token, appends it to the input, and repeats. Instruction-following and reasoning emerge from this objective plus **post-training** (instruction tuning, preference training). prompt tokens The cat sat on transformer <text cla

1.1k views

Understand the problem

The transformer, next-token prediction, and why 'autocomplete at scale' produces reasoning-like behaviour.

llmtransformerfundamentalsnext-token

Attempt it yourself

Sketch your approach before reading the solution — that's what interviews test.

Stuck? AI Nudge Available

Get a conceptual hint to guide your logic without spoiling the final implementation.

Study the solution

The solution is waiting

Give it an honest attempt first — then compare your thinking with the full walkthrough.

Read the code

One decoding step, conceptually

context = tokenize("The cat sat on")
while not done:
    logits = transformer(context)          # scores for every vocab token
    probs  = softmax(logits[-1] / temperature)
    tok    = sample(probs)                 # greedy / top-p / top-k
    context.append(tok)                    # feed it back in
    done   = (tok == EOS) or len(context) >= max_tokens

Join the discussion

Discussion (0)

No responses yet. Be the first to share what you think.

What are embeddings and what are they used for?