Proj. CLJ Paper Reading: Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrations

Abstract

• Tool: In-Context Learning
  • Tool1: In-Context Attack
  • Tool2: In-Context Defense
• Task: modulate the alignment of LLMs, especially Safety alignment

1. intro

P1: 什么是jailbreak

P2: Categories of jailbreak

• Optimization based: 效率瓶颈
• Template based: 缺乏可拓展性和灵活性
  • 本文，利用in-context demonstration, 灵活可拓展