MAPRO: Recasting Multi-Agent Prompt Optimization as Maximum a Posteriori Inference

Open in new window