This survey introduces a new field called Context Engineering, which is about carefully choosing and organizing the information we give to Large Language Models (LLMs), like ChatGPT, so they can work better. It’s more than just writing good prompts — it’s about designing the full package of information the model uses to think and respond.
在我还在读小学的时候，唯一开过的网络服务就是QQ会员了。红色的昵称显示，在好友列表里优先显示，还有形形色色的聊天气泡，主页空间装扮。不过那时候也没钱，用的都是各种网上找的bug开上的。不过也算是第一次的订阅经历吧。
October 28 Monday 2024, there are 54 days left before the Postgraduate Entrance Examination. 9:30 am, I wake up from bed. I have been slacking off quite a lot—sleeping late, not waking up on time, and completing fewer tasks than I supposed to do. For this, I’ve conclude the following reasons.
HE bent down and rubbed his ankle where it had been sprained with the walking so that he couldn’t see the ankle bone. Then he straightened up and looked around him. He felt in his pocket for a packet of cigarettes, took one out and lit it. He wiped the sweat from his forehead with the back of his hand and he stood in the middle of the street looking around him.
能够写论文的致谢部分，说明我已经完成了论文，回望这段经历，百分之八十的时间都是在漫无目的的追着什么，像是一个个矩阵变换叠加的向量那无序的乱舞，像是反常积分看不到收敛。直到真正看见什么值得追的东西，才算是矩阵变换求出了特征向量，黎曼积分的和有了正值。
元旦家里来了小孩，感觉就像来个了小祖宗啊。服了，爸爸妈妈叫我陪他玩。我拿出我的switch，想看他玩塞尔达。因为看新人玩塞尔达真的很有趣，结果他没有一点兴趣。
现在回过头来看，真的很神奇。一开始我甚至是连神奇树屋都看不懂的。那些简单的单词，组成的句子我竟然看不懂。以前只记得每个单词的意思，做题的话在脑海里翻译成中文就能做了。现在看书上这些口语话的表达这招就不管用了，真的是单纯认识却不知道意思了，更何况这些句子都是一些简单句，也不存在语法障碍。看来是真的缺少native那种儿童口语化英语英文输入了。
这是很难想象的事，大学生活就这么快结束了。这篇文章本来是学校要求写的给学弟学妹的建议，但对这样一个没什么并不成功的我来说，我想我是没什么资格建议的。
一本自己的词书，相遇来自考研词汇
早上起床，我想也不能称之为早上，发现昨天开始的症状并没有减轻。昨天早上开始，屁股眼里的肌肉便开始不受控制的的筋挛，好像在昭示自己的存在般。虽不严重，但总让我心烦，完全没法沉浸在眼前的学习中。
Update your browser to view this website correctly. Update my browser now