AUX
GitHub - 1999labs/monkey-see-monkey-do: Two mechanically-scored evals for LLMs: rule inference from examples, and derivation in a novel string-rewrite system. — AUX