Paper-Conference

Prompt Optimization via Adversarial In-Context Learning
We propose a new method, Adversarial In-Context Learning (adv-ICL), to optimize prompts for in-context learning (ICL). Inspired by …
PDF Cite