Learning language by compression — an MDL language learner with no transformer and no backprop. Predicts, chunks, induces grammatical categories, and learns new words in one shot. - View it on GitHub