用视觉语言模型做外部安全审计，防后门攻击更可靠。
From Internal Diagnosis to External Auditing: A VLM-Driven Paradigm for Data-Free Online Backdoor Defense
- 用通用视觉语言模型当独立审计员，不依赖受损模型
- 在线动态更新视觉原型，实时调节防御阈值
- 17个数据集11种攻击下成功率低于1%，适合部署防护
深度神经网络仍易受后门攻击。传统测试时防御多依赖模型修复或输入鲁棒性等内部诊断方法，但在高级攻击下常因与受损模型参数耦合而失效。本文提出从内部诊断转向外部语义审计的新范式，主张通过独立的、语义基础的审计器解耦安全性。为此，我们设计基于通用视觉语言模型（VLM）的框架，提出PRISM（原型精炼与统计监控检测），通过混合VLM教师在线动态优化视觉原型，以及基于统计置信度监控的自适应路由机制，实时校准决策阈值。在17个数据集和11类攻击下的实验表明，PRISM在CIFAR-10上将攻击成功率降至<1%，同时保持高干净准确率，确立了无需模型信息、可外部化部署的新型安全标准。
原文摘要 · Abstract (English)
Deep Neural Networks remain inherently vulnerable to backdoor attacks. Traditional test-time defenses largely operate under the paradigm of internal diagnosis methods like model repairing or input robustness, yet these approaches are often fragile under advanced attacks as they remain entangled with the victim model's corrupted parameters. We propose a paradigm shift from Internal Diagnosis to External Semantic Auditing, arguing that effective defense requires decoupling safety from the victim model via an independent, semantically grounded auditor. To this end, we present a framework harnessing Universal Vision-Language Models (VLMs) as evolving semantic gatekeepers. We introduce PRISM (Prototype Refinement & Inspection via Statistical Monitoring), which overcomes the domain gap of general VLMs through two key mechanisms: a Hybrid VLM Teacher that dynamically refines visual prototypes online, and an Adaptive Router powered by statistical margin monitoring to calibrate gating thresholds in real-time. Extensive evaluation across 17 datasets and 11 attack types demonstrates that PRISM achieves state-of-the-art performance, suppressing Attack Success Rate to <1% on CIFAR-10 while improving clean accuracy, establishing a new standard for model-agnostic, externalized security.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。