Attacks, Defenses and Evaluations for LLM Conversation Safety: A Survey

TL;DR

This survey comprehensively reviews recent research on attacks, defenses, and evaluation methods for ensuring safety in Large Language Model conversations, highlighting current challenges and future directions.

Contribution

It provides a structured taxonomy of studies on LLM conversation safety, summarizing recent advances and identifying gaps for future research.

Findings

Categorized existing studies into attacks, defenses, and evaluations.

Highlighted key challenges in LLM conversation safety.

Provided a structured overview to guide future research efforts.

Abstract

Large Language Models (LLMs) are now commonplace in conversation applications. However, their risks of misuse for generating harmful responses have raised serious societal concerns and spurred recent research on LLM conversation safety. Therefore, in this survey, we provide a comprehensive overview of recent studies, covering three critical aspects of LLM conversation safety: attacks, defenses, and evaluations. Our goal is to provide a structured summary that enhances understanding of LLM conversation safety and encourages further investigation into this important subject. For easy reference, we have categorized all the studies mentioned in this survey according to our taxonomy, available at: https://github.com/niconi19/LLM-conversation-safety.