← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6606.b · 生成模型:GPT-6

零样本提示明确输出空间

理解等级:中级

零样本提示不提供任务示例,但可明确指令、输入与允许标签或格式。像分类任务可先限定输出集合,再在代表样本上检查边界;这不能保证模型总能遵循格式,自动系统仍需验证输出。

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

      Describe the output space or the range of acceptable outputs and output format. In this example, we state the output should be one of three values. Prime it to generate the correct text. By ending the prompt with “Sentiment:,” we are increasing the probability of the LLM generating the sentiment value as the next token. The better the model, the less you have to worry

    参考来源

    - 位置:《Designing Large Language Model Applications》第01章《大语言模型入门》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch01.html) - 结论依据:摘录中关于零样本提示明确输出空间的定义、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:Describe the output space or the range of acceptable outputs and output format. In this example, we state the output should be one of three values. Prime it to generate the correct text. By ending the prompt with “Sentiment:,” we are increasing the probability of the LLM generating the sentiment value as the next token. The better the model, the less you have to worry