English(EN) SEMA-Auto: Sighted and Efficient Multi-Turn Attack with Autonomy

新的SEMA-Auto系统在大型语言模型上实现了最先进的越狱

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-27 04:00

研究人员开发了SEMA-Auto，一个旨在绕过大型语言模型安全措施的新型自主系统。与预先规划攻击或需要昂贵交互式评估的先前方法不同，SEMA-Auto可以独立评估其成功并终止会话。该系统采用两阶段训练过程，包括合成引导调优和决策监督强化学习，以提高其有效性和效率。SEMA-Auto在包括GPT-oss-120B和Claude-Haiku-4.5等前沿模型在内的各种数据集和模型上，展示了最先进的攻击成功率，优于现有的静态和交互式基线。 AI

影响这项研究突显了大型语言模型安全机制的潜在漏洞，并可能促进更强大防御措施的开发。

排序理由研究论文，详细介绍了一种攻击大型语言模型的新方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 10 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 研究论文，详细介绍了一种攻击大型语言模型的新方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: safety, paper; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Mingqian Feng, Xiaodong Liu, Weiwei Yang, Chenliang Xu, Jialin Song, Qiyao Ma, Chengquan Guo, Carolyn Buractaon, Christopher White, Jianfeng Gao · 2026-10-27 04:00

SEMA-Auto：具有自主性的可见且高效的多轮攻击

arXiv:2610.35743v1 Announce Type: new Abstract: Existing jailbreak attackers fall into two classes: static planners fix their single- or multi-turn adversarial plans in advance, remaining simple but losing sight of the victim's replies, wasting turns and limiting transferability;…

本文如何被排名

报道来源 [1]

SEMA-Auto：具有自主性的可见且高效的多轮攻击

相关实体

相关话题