← Back to feed

Agent Anomaly Detection scores on benchmarks we didn't write

AI · Security · 2026-11-13 16:33

谷歌在第三方公开基准上测试其智能体异常检测功能，验证其在识别目标劫持等安全风险方面的表现。

Source: developers.googleblog.com

Auto-generated summary for reference only. RSS