@potpie @plexus
If you have a sentence, an LLM can assign a "probability of coming next" to each word in its vocabulary. For someone interested in the statistics of language, that is a *kind* of model...