paper-with-me

홈 › Papers

OTTER: A Red-Teaming System for Toxicity-Evading Jailbreak Prompt Optimization

2026-06-19 · Jerry Wang, Hsin-Ling Hsu, Yi-Cheng Lai, Nai-Chia Chen, Fang Yu arxiv

Production LLMs increasingly rely on toxicity-based moderation filters as a primary defense, assuming that harmful intent correlates with toxic surface wording. We show this assumption is fundamentally brittle: surface toxicity and adversarial intent can be decoupled by replacing as few as five tokens. We present OTTER (Obfuscated Toxicity-Evading Token Evolution for Rewriting), a black-box red-teaming framework requiring only standard API access, directly targeting the practical constraints of industry security audits. Evaluated on 457 AdvBench prompts across four GPT models, OTTER raises average ASR from 7.0% to 84.0%. We further provide the first quantitative analysis of the toxicity--bypass relationship and a per-category breakdown, translating our findings into actionable recommendations for classifier hardening in production deployments.

📄 PDF Abstract BibTeX arXiv:2606.21077

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Reverse Constitutional AI: A Framework for Controllable Toxic Data Generation via Probability-Clamped RLAIF

2026-04-20 · Yuan Fang, Yiming Luo, Aimin Zhou, Fei Tan arxiv

Ensuring the safety of large language models (LLMs) requires robust red teaming, yet the systematic synthesis of high-quality toxic data remains under-explored. We propose Reverse Constitutional AI (R-CAI), a framework f…

Reinforcement LearningRed Teaming

WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models

2024-06-26 · Liwei Jiang, Kavel Rao, Seungju Han, Allyson Ettinger 외

We introduce WildTeaming, an automatic LLM safety red-teaming framework that mines in-the-wild user-chatbot interactions to discover 5.7K unique clusters of novel jailbreak tactics, and then composes multiple tactics for…

ChatbotRed Teaming

POEX: Understanding and Mitigating Policy Executable Jailbreak Attacks against Embodied AI

2024-12-21 · Xuancun Lu, Zhengxian Huang, Xinfeng Li, Xiaoyu Ji 외

Embodied AI systems are rapidly evolving due to the integration of LLMs as planning modules, which transform complex instructions into executable policies. However, LLMs are vulnerable to jailbreak attacks, which can gen…

LLM JailbreakRed Teaming

RedAgent: Red Teaming Large Language Models with Context-aware Autonomous Language Agent

2024-07-23 · Huiyu Xu, Wenhui Zhang, Zhibo Wang, Feng Xiao 외

Recently, advanced Large Language Models (LLMs) such as GPT-4 have been integrated into many real-world applications like Code Copilot. These applications have significantly expanded the attack surface of LLMs, exposing …

Red Teaming

ART: Automatic Red-teaming for Text-to-Image Models to Protect Benign Users

2024-05-24 · Guanlin Li, Kangjie Chen, Shudong Zhang, Jie Zhang 외

Large-scale pre-trained generative models are taking the world by storm, due to their abilities in generating creative content. Meanwhile, safeguards for these generative models are developed, to protect users' rights an…

DiversityLanguage ModelingLanguage ModellingLarge Language Model+1