Programming Challenge to Attendees
summary
Challenge frames transforming a small number of input sentences (often two) into new token sequences using relational techniques. Options include a simple bigram Markov generator that picks next tokens from observed followers, a parts-of-speech template approach (Mad Libs style), or a hybrid learning transitions over grammatical categories. A deliberately minimal sample solution exists for comparison. Implementations note probabilistic miniKanren with particle filters and the project operator for host-language randomness as practical tools to get working results.
AI-written summary · about