知识卡片

普通读书笔记卡 · 6784.d · 生成模型:GPT-6

状态节点不等于全部运行上下文

理解等级:中级

图中节点标识当前步骤,运行实例还包含槽位、历史、用户、重试数等状态。多个用户可处于同一节点却有不同上下文;把图定义与会话实例分开,才能防止串话,并支持独立恢复。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Imagine you want to represent a math tutoring conversation as a graph. What would the nodes of the graph represent? What about the edges? Different chatbot platforms treat this question differently. Most platforms provide you with some building blocks (components) that you can combine to create the conversation graph. But at its core, inside the chatbot software itself, the nodes are the chatbot’s state—everything the

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第12章《对话引擎与会话设计》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-12.html) - 结论依据:摘录中关于状态节点不等于全部运行上下文的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Imagine you want to represent a math tutoring conversation as a graph. What would the nodes of the graph represent? What about the edges? Different chatbot platforms treat this question differently. Most platforms provide you with some building blocks (components) that you can combine to create the conversation graph. But at its core, inside the chatbot software itself, the nodes are the chatbot’s state—everything the