Jev 与文本生成模型的取舍
Jev 不生成自由文本。Diogo 用一段对比视频说明并行决策模型与逐字生成模型在任务形态上的差别。

帖子指出,提示注入之所以有效,是因为模型能写出下一条指令;而 Jev 的结构化评估模型只输出类型化的概率、标签或刻度值,恶意文本只能影响数值,无法变成命令。
Prompt injection works because the model can write the next instruction. A structured evaluation model cannot. Jev's only output is typed a probability, a label, a point on a scale. Text that tries to hijack an agent can move the number. It cannot become a command. That is why