技术英语精读 · 句式篇
📖 Controlling Reasoning Effort in LLMs（控制大语言模型的推理强度） 作者：Sebastian Raschka｜原文：https://magazine.sebastianraschka.com/p/controlling-reasoning-effort-in-llms
✍️ 句式拆解：
“It has been almost two years since OpenAI released o1, a model that popularized the idea of LLM-based reasoning models.”
🔍 结构：
- 主句：
It has been almost two years— 现在完成时 “It has been + 时间段”，表达”从……至今已有……” - 时间状语从句：
since OpenAI released o1— since 引导，点明时间起点 - 同位语：
a model— 对o1作补充说明 - 定语从句：
that popularized the idea of LLM-based reasoning models— 修饰a model，进一步解释其历史意义
💡 亮点：
- 一句三层：先交代时间跨度（It has been…since…），再用「同位语 + 定语从句」把”o1 是什么、有何意义”打包进一个名词短语，信息密度高却不臃肿。
It has been + 时间 + since + 从句是科普/新闻写作的开篇高频句式，天然适合定调。popularized用词精准：不是”发明”（invented），而是”普及/推广”，体现作者对技术史的准确判断——o1 并非推理模型的开创者，而是把它带进大众视野。
🧩 段落精读：
“Notably, the reasoning trace itself was not used for training or updating the model. Although they tried to use this intermediate response information for training, the DeepSeek-R1 paper reported that it wasn’t helpful for the model training, so it was ultimately not used. (Whether and how to incorporate intermediate reasoning traces in the training signal via process reward models is an active area of research.)”
🔗 逻辑：
- 论点（第 1 句）：推理轨迹本身并未被用于训练——用
Notably引出反直觉的结论。 - 论证（第 2 句）：
Although让步 +so因果，说明”试过但没用，最终放弃”。 - 结论/留白（第 3 句）：括号补充，把话题引向更前沿的开放问题，为后文埋下伏笔。
📌 连接词：Notably / Although / so / ultimately
🎯 模仿练习：用类似结构写一句技术表达
“It has been over a decade since React introduced the virtual DOM, a concept that reshaped how frontend engineers think about rendering.”
（套用 “It has been…since…, a concept that…” 结构，一句交代时间、对象与意义。）
📎 备注：原文由 Sebastian Raschka 于 2026-07 发布，系统讲解推理模型的训练流程与”推理强度（reasoning effort）“控制机制。