arXiv:2608. 16465v1 Announce Type: new Abstract: Automated red-teaming has produced a growing collection of attack strategies, yet they typically remain scattered across prompts and workflows, making them difficult to systematically integrate, reuse, and improve at scale.
By Xiaoyu Wen, Jiajia Li, Zhida He, Peng Yu, Chenxu Wang, Han Qi, Ziyuan Zhou, Cheng Jin, Ying Wen, Xingcheng Xu, Shuyue Hu, Tianhang Zheng, Chaochao Lu, Qiaosheng Zhang
EvoFlint is an evolutionary atlas that maps multi‑turn vulnerabilities in large language models by treating attack discovery as a search problem rather than a generation task. It uses evolutionary quality‑diversity search to evolve phased conversation plans, employing Pareto fitness for success rate and severity, novelty search for diversity, and a generation‑level memory to incorporate model insights. The resulting risk‑indexed archive, tested on HarmBench, shows high attack success rates across models such as Claude Sonnet, GPT‑5, and Qwen3, revealing which harm categories each model’s safety training covers or misses.
By Feitong Qiao, Liren Peng, Shiming Ren, Aishwarya Jadhav, Arghavan Bahadorinejad, Marinette Chen, Muhan Zhang, Abdulaziz Suria, Gennevi Lu, Anish Das Sarma
RedEvoAgent is a black-box red‑teaming agent that transforms cross‑case attack trajectories into concise, human‑readable attack skills. It evolves these skills by profiling tool effectiveness, attributing tool credit, and applying a validation ratchet to keep only improvements. Experiments demonstrate that RedEvoAgent outperforms fixed and agentic baselines, enhances tool efficiency, and transfers across attacker models and target execution harnesses.
By Junjie Zhang, Hui Liu, Kecheng Chen, Xianbo Mo, Changsheng Chen, Haoliang Li
The paper introduces ADVERSA, an automated red‑teaming framework that evaluates large language model safety over multiple turns by tracking continuous compliance trajectories instead of binary jailbreak outcomes. Using a fine‑tuned 70B attacker model and a structured 5‑point rubric, the authors conduct controlled experiments on three frontier victim models, measuring guardrail degradation and judge reliability through a triple‑judge consensus. Results show a 26.7% jailbreak rate with most breaches occurring early, and the study documents inter‑judge agreement, attacker drift, and attacker refusals as key factors affecting safety assessment.
By Harry Owiredu-Ashley
The paper examines the vulnerability of open‑weight large language model safeguards that aim to block harmful usage by assuming new harmful behavior arises only through fine‑tuning. It demonstrates that simple, low‑cost attacks—abliteration and prefilling—can significantly increase attack success rates across multiple benchmarks, and proposes an abliteration‑resistant tuning (ART) method that mitigates these attacks. The study shows that safeguarding defenses must be evaluated against a broader range of attack strategies beyond fine‑tuning.
By Kevin Kuo, Virginia Smith, Chhavi Yadav
arXiv:2506. 07121v2 Announce Type: replace 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.
By Ren-Jian Wang, Ke Xue, Zeyu Qin, Ziniu Li, Sheng Tang, Hao-Tian Li, Shengcai Liu, Zhi Yu, Yuanpeng Tan, Chao Qian