文章警告，经过修改的开放权重模型（如去拒绝化版本）可被植入后门，能通过评测且几乎躲过现有所有检测，直到隐藏触发条件激活。
被植入后门的模型能通过评测，随你的智能体一起部署，正常回答每一个请求，然后静静等待。触发条件由修改者任意设定：一个罕见短语、一个客户名字、一个尚未到来的日期。如果你技术栈中的某个模型被这样植入后门，你现在运行的几乎任何工具都不会察觉。原文 (en)
A backdoored model passes the evals, ships inside your agent and answers every normal request, then waits. The trigger is whatever the person who edited it chose, a rare phrase, a customer’s name, a date that hasn’t come yet. If a model in your stack were backdoored this way, almost nothing you run today would notice.