How LLMs work, part 3: attention, queries, keys and values

Attention is the one idea every large language model is built on. It lets each token look at every other token in the text and decide which ones matter for understanding it. This is part 3 of our series on how LLMs work. You will see why attention replaced reading word by word, how it resolves meaning in a real sentence, and the exact math of queries, keys and values, without skipping steps.