Explore

Trigram Language Model

Here we explore the conditional probability distributions for a trigram language model trained on 30,000 names. This is a character-level model with 0.1 smoothing. The heatmap below shows the probabilities of the next character given the previous two characters. Roll over the heatmap (or tap on mobile) to see the probabilities. Choose two characters by selecting from the dropdowns below to condition the heatmap on those characters and focus on the probabilities of the next character given those two characters. (The \n character is shown as ↵)