创建

登录 ReadmeX

或

了解我们如何处理你的信息: 隐私政策

ReadmeX

陪你读懂 AI 动态

今天,想先了解什么?

选一份动态总结,或直接聊聊你的问题。

社区和关注动态总结只属于你自己。

资讯

Anthropic称AI安全分数高度依赖测试设计

AI 总结

Anthropic发布可靠AI安全测试指南,认为模型是否更安全取决于其背后的测试是否可靠。其指南称,Claude Fable 5.1在API自残测试对话中的适当行为比例为60%,而在加入隐藏指令的Claude应用中为94%。Anthropic还建议采用更长的对话、由人类专家校验AI评分器,并公开结果的不确定性。

深度见解

据Anthropic自身数据,同一模型在API与应用中的结果相差34个百分点,显示提示词和测试设置会显著改变所报告的安全表现。

Anthropic says its AI safety scores depend heavily on how they’re measured
Anthropic says the same model can score very differently on safety tests depending on how they are set up. Credit: Anthropic
Image Credits
Anthropic says the same model can score very differently on safety tests depending on how they are set up. Credit: Anthropic

Anthropic has published advice on building AI safety tests. It shows Claude Fable 5.1 behaved appropriately in only 60% of self-harm test conversations via its API, against 94% in the Claude app after a fix, and warns that many headline safety scores hide wide uncertainty.

Anthropic has published advice on how to build reliable AI safety tests, and it includes some unflattering numbers about its own models. The guide, out on Thursday, says a claim that one model is “safer” is only as good as the test behind it.

#Anthropic#Claude Fable 5.1#Claude Opus 5#Claude Opus 5.5

原文约 2 分钟读完阅读原文
群聊

还没有评论,来说说你的看法。