A series of LLM experiments powered by the Swiss Apertus-v1.5-70b model, with inference provided by the Public AI Inference Utility.