https://terms-en.ai-term-hub.com/zh/tags/ai-security/Recent content in AI Security on 中文AI术语词典Hugozh-cnSat, 18 Jul 2026 11:44:45 +0000https://terms-en.ai-term-hub.com/zh/terms/backdoor_attack/Sat, 18 Jul 2026 11:38:27 +0000https://terms-en.ai-term-hub.com/zh/terms/backdoor_attack/<h2 id="definition">Definition</h2> <p>后门攻击涉及使用特定的模式（称为触发器）对机器学习模型的训练数据进行投毒。虽然模型在干净数据上表现正常，但在检测到触发器时会激活错误的行为，例如将特定图像错误分类为指定标签，</p> <h3 id="summary">Summary</h3> <p>一种安全威胁，恶意行为者在训练期间将隐藏触发器嵌入 AI 模型，导致特定错误分类。</p> <h2 id="key-concepts">Key Concepts</h2> <ul> <li>数据投毒</li> <li>模型完整性</li> <li>对抗性机器学习</li> <li>触发模式</li> </ul> <h2 id="use-cases">Use Cases</h2> <ul> <li>ML 流水线的安全测试</li> <li>开发防御过滤器</li> <li>审计第三方模型</li> </ul> <h2 id="related-terms">Related Terms</h2> <ul> <li><a href="https://terms-en.ai-term-hub.com/en/terms/adversarial-examples-%E5%AF%B9%E6%8A%97%E6%A0%B7%E6%9C%AC/">Adversarial Examples (对抗样本)</a></li> <li><a href="https://terms-en.ai-term-hub.com/en/terms/data-poisoning-%E6%95%B0%E6%8D%AE%E6%8A%95%E6%AF%92/">Data Poisoning (数据投毒)</a></li> <li><a href="https://terms-en.ai-term-hub.com/en/terms/model-robustness-%E6%A8%A1%E5%9E%8B%E9%B2%81%E6%A3%92%E6%80%A7/">Model Robustness (模型鲁棒性)</a></li> <li><a href="https://terms-en.ai-term-hub.com/en/terms/clean-label-attacks-%E5%B9%B2%E5%87%80%E6%A0%87%E7%AD%BE%E6%94%BB%E5%87%BB/">Clean Label Attacks (干净标签攻击)</a></li> </ul>