针对SmoothLLM防御框架提出概率性认证方法，突破传统确定性保证局限，为LLM对抗鲁棒性提供更贴近实际场景的量化评估。
arXiv:2511.18721v4 Announce Type: replace-cross Abstract: The SmoothLLM defense provides a certification guarantee against jailbreaking attacks, but i…
首个聚焦安全关键控制室的LLM操作员多轮红队基准，覆盖对抗鲁棒性与越狱攻击评测。
arXiv:2606.20408v1 Announce Type: cross Abstract: Large language model (LLM) agents are increasingly proposed as supervisory components for safety-cri…
无需训练即可增强计算MRI的对抗鲁棒性，ICML 2026论文提出全新方法。
arXiv:2501.01908v4 Announce Type: replace-cross Abstract: Deep learning (DL) methods have become the state-of-the-art for reconstructing sub-sampled m…
多模态大语言模型充当评判者是否可靠？本文揭示其应对对抗攻击时的脆弱性并提出提升鲁棒性的方向。
arXiv:2606.15608v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) are increasingly used as automated judges, e.g., for image qu…
最新研究揭示：测试时训练（TTT）能绕过AI安全护栏，引发对模型防御机制的新思考。
arXiv:2605.22984v1 Announce Type: cross Abstract: Test-Time Training (TTT) is an emerging paradigm that enables models to adapt their parameters durin…
针对大模型安全对齐中上下文敏感漏洞，提出创新方法实现跨场景一致性防护。
arXiv:2605.20994v1 Announce Type: new Abstract: Preference-based post-training aligns LLMs with human intent, yet safety behavior often remains brittl…
首个专注多智能体LLM集体对抗鲁棒性的三模式基准，揭示单一欺骗智能体如何突破现有防御。
arXiv:2605.09027v2 Announce Type: cross Abstract: In multi-agent systems (MAS), a single deceptive agent can nullify all gains of an agentic AI collec…