arXiv cs.AI· Ren-Jian Wang, Ke Xue, Zeyu Qin, Ziniu Li, Sheng Tang, Hao-Tian Li, Shengcai Liu, Zhi Yu, Yuanpeng Tan, Chao Qian·· 13 小时前AI 评分38
QDRT：为大语言模型自动生成高质量多样化攻击者的质量-多样性红队框架
Quality-Diversity Red-Teaming: Automated Generation of High-Quality and Diverse Attackers for Large Language Models
AI 导读
研究者提出质量-多样性红队框架 QDRT，通过行为条件训练与开放式行为回放缓冲区实现目标驱动的攻击多样性，并训练多个专门化攻击者覆盖不同攻击风格与风险类别。评估显示，QDRT 生成的攻击在 GPT-2、Llama-3、Gemma-2、Qwen2.5 及商用模型 GPT-4.1、GPT-5-Chat 上兼具更强多样性与攻击效果。代码已开源。
正文 · 原文
Abstract:Ensuring the safety and robustness of large language models (LLMs) is a fundamental challenge and a critical prerequisite for the responsible deployment of artificial intelligence. Red-teaming, a systematic framework to identify adversarial prompts that elicit harmful responses from target LLMs, has emerged as a crucial safety evaluation paradigm. Within this framework, the diversity of adversarial prompts is critical for a comprehensive safety assessment. However, previous red-teaming approaches often pursue diversity through simplistic metrics such as word frequency or sentence embedding similarity, which may not capture meaningful variation in attack strategies. In addition, the common practice of training a single attacker model restricts coverage across all potential attack styles and risk categories. This paper introduces Quality-Diversity Red-Teaming (QDRT), a new framework designed to address these limitations. QDRT achieves goal-driven diversity through behavior-conditioned training and implements a behavioral replay buffer in an open-ended manner. In addition, it trains multiple specialized attackers capable of generating high-quality attacks across diverse styles and risk categories. Our empirical evaluation demonstrates that QDRT generates attacks that are both more diverse and more effective against a wide range of target LLMs, including open-source models \texttt{GPT-2}, \texttt{Llama-3}, \texttt{Gemma-2}, \texttt{Qwen2.5}, and commercial models \texttt{GPT-4.1} and \texttt{GPT-5-Chat}. This work advances the field of LLM safety by providing a systematic and effective approach to automated red-teaming, ultimately supporting the responsible deployment of LLMs. Our code is available at this https URL.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE) |
| Cite as: | arXiv:2506.07121 [cs.LG] |
| (or arXiv:2506.07121v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2506.07121 arXiv-issued DOI via DataCite |
Submission history
From: Chao Qian [view email]
[v1]
Sun, 8 Jun 2025 13:07:41 UTC (405 KB)
[v2]
Sat, 1 Aug 2026 10:21:24 UTC (465 KB)
[v3]
Wed, 21 Oct 2026 07:48:18 UTC (443 KB)
来源：arXiv cs.AI · arxiv.org