按年被提到
20242024 · 2 次2026
AI 安全、伦理与社会影响›模型窃取、记忆与水印·arXiv:2412.03556 · cs.CL

Best-of-N Jailbreaking

John Hughes, Sara Price, Aengus Lynch, Rylan Schaeffer, Fazl Barez, Sanmi Koyejo, Henry Sleight, Erik Jones, Ethan Perez, Mrinank Sharma · 2024
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 3 人首次 2024-12-13 · 最近 2025-02-20
3不同的人
2活跃月份
3讨论串
24被引(OpenAlex)
arXiv 摘要(候选层不写导读)

We introduce Best-of-N (BoN) Jailbreaking, a simple black-box algorithm that jailbreaks frontier AI systems across modalities. BoN Jailbreaking works by repeatedly sampling variations of a prompt with a combination of augmentations - such as random shuffling or capitalization for textual prompts - until a harmful response is elicited. We find that BoN Jailbreaking achieves high attack success rates (ASRs) on closed-source language models, such as 89% on GPT-4o and 78% on Claude 3.5 Sonnet when sampling 10,000 augmented prompts. Further, it is similarly effective at circumventing state-of-the-art open-source defenses like circuit breakers. BoN also seamlessly extends to other modalities: it jailbreaks vision language models (VLMs) such as GPT-4o and audio language models (ALMs) like Gemini 1.5 Pro, using modality-specific augmentations. BoN reliably improves when we sample more augmented prompts. Across all modalities, ASR, as a function of the number of samples (N), empirically follows power-law-like behavior for many orders of magnitude. BoN Jailbreaking can also be composed with other black-box algorithms for even more effective attacks - combining BoN with an optimized prefix attack achieves up to a 35% increase in ASR. Overall, our work indicates that, despite their capability, language models are sensitive to seemingly innocuous changes to inputs, which attackers can exploit across modalities.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论chrisshawyeah · 2025-02

“Hi HN! I built an open-sourced best-of-n-prompt-jailbreaker tool https://bestofn.vetonce.com/ since I heard about this technique and haven't seen any updates about it since https://arxiv.org/abs/2412.03556 Inspired by https://github.com/jplhughes/bon-jailbreaking Would love your feedback”

Best-of-n-prompt-jailbreaker – open-source AI prompt tool原评论 ↗
被提交为 story · 2 次