标签: OpenAI

Suggested patches open in Claude Code on the web, using the models your team already uses. These updates are part of our work to give defenders greater access to Mythos results without requiring direct access to the m…

Suggested patches open in Claude Code on the web, using the models your team already uses. These updates are part of our work to give defenders greater access to Mythos results without requiring direct access to the m...

Anthropic 将 Claude Security 扫描功能升级至 Mythos 5 模型,并向所有 Claude Enterprise 客户开放公开测试版;该模型运行在扫描后台,企业无需单独申请模型访问权限,即可获得漏洞检测与修复建议。

debating whether we should make it an official plugin, but you can try installing it here to try with these commands: claude plugin marketplace add anthropics/claude-plugins-community claude plugin install eli5@claude…

debating whether we should make it an official plugin, but you can try installing it here to try with these commands: claude plugin marketplace add anthropics/claude-plugins-community claude plugin install eli5@claude...

Anthropic 内部员工常用的 ELI5("Explain Like I'm 5")技能正被推向外部社区,用户可通过命令行安装体验。这个功能本质上是让 AI 用图文并茂的 HTML 页面,把复杂概念讲给零基础的人听。

Gemini for Students

Gemini for Students

谷歌面向学生群体推出 Gemini 专用入口,将大模型对话、资料检索与学习辅助场景打包在一起,试图在校园场景建立 AI 工具的入口优势。

心理学方法揭示AI安全测试的重大弱点

心理学方法揭示AI安全测试的重大弱点

一项由英国 AI 安全研究所等机构参与的大规模研究显示,主流大模型安全基准测试的“综合安全分”很容易被刷高——模型只要无差别拒绝更多请求就能得分,但这会让模型变得难用。研究还发现,绝大多数测试题是冗余的,并首次用统计学方法识别出“装乖”的模型。