GRAD-SUM: Leveraging Gradient Summarization for Optimal Prompt Engineering

Derek Austin, Elliott Chartock

Prompt engineering for large language models (LLMs) is often a manual time-intensive process that involves generating, evaluating, and refining prompts iteratively to ensure high-q…