知识卡片

普通读书笔记卡 · 6637.a · 生成模型:GPT-6

领域记号改变学习表示成本

理解等级:中级

若任务本身有紧凑记号,可用领域词表减少表达冗余,让模型更直接接触操作结构。但记号设计仍须保留任务语义与可解析性;表达更短不保证模型自动遵守领域规则。

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

      Note that this model was able to learn the concepts (chess) using a domain-specific language (PGN). How will we fare if the concepts were taught in natural language? Let’s explore this in another experiment. Take the same dataset used to pre-train the chess language model and run it through an LLM to convert each move in PGN to a sentence in English. An example game

    参考来源

    - 位置:《Designing Large Language Model Applications》第04章《架构与学习目标》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch04.html) - 结论依据:摘录中关于领域记号改变学习表示成本的定义、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:Note that this model was able to learn the concepts (chess) using a domain-specific language (PGN). How will we fare if the concepts were taught in natural language? Let’s explore this in another experiment. Take the same dataset used to pre-train the chess language model and run it through an LLM to convert each move in PGN to a sentence in English. An example game