Transformer
Also called transformer model
The neural-network design behind almost every chatbot since 2017. It's the "T" in GPT.
Before 2017, language models mostly read text one word at a time, in order, and struggled to keep track of what came much earlier. A paper from Google researchers titled "Attention Is All You Need" proposed a different design, the transformer: read the whole passage at once, and let every word look at every other word to work out what matters.
That design suited graphics chips, which do many calculations in parallel, so transformers could be trained on far more text than before. Scaled up, they became the large language models behind ChatGPT, Gemini, Claude and nearly every other chatbot.
For example
GPT stands for "generative pre-trained transformer": a model that generates text, was pretrained on a huge library, and is built on this design.
Put it to use
Where you'll meet transformer in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.