https://guijiagi.com/tags/%E9%98%B2%E5%BE%A1/Recent content in 防御 on 硅基 AGI · 智能体学习与测评Hugozh-cn本站内容采用 CC BY-NC-SA 4.0 国际许可协议授权Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/agent-jailbreak-defense/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/agent-jailbreak-defense/AI Agent越狱攻击防御https://guijiagi.com/posts/prompt-injection-defense-panorama/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/prompt-injection-defense-panorama/Prompt注入攻击全景防御https://guijiagi.com/posts/prompt-injection-defense/Fri, 26 Jun 2026 14:00:00 +0800https://guijiagi.com/posts/prompt-injection-defense/系统讲解 Prompt 注入攻击的原理、分类与防御策略，包含红队测试方法、代码实现和生产级防御架构。https://guijiagi.com/posts/model-stealing-defense/Thu, 25 Jun 2026 10:00:00 +0800https://guijiagi.com/posts/model-stealing-defense/全面解析 LLM 模型窃取攻击技术（API 提取、知识蒸馏、侧信道）及防御策略、检测方法和法律保护手段。