Joint Prompt Optimization of Stacked LLMs using Variational Inference

Open in new window