内容来源：https://blog.google/innovation-and-ai/technology/developers-tools/colab-updates/

内容总结：

谷歌Colab平台近日推出两项基于Gemini模型的新功能——“自定义指令”与“学习模式”，旨在为用户打造个性化编程学习助手。

“自定义指令”功能允许用户在笔记本层级设置个性化偏好，例如指定代码风格、关联课程大纲或优先使用特定编程库等。这些指令将随笔记本保存，分享给社区协作时，定制化AI体验也能同步传递。

“学习模式”则将Gemini转变为编程导师，通过分解复杂概念、提供分步指导来引导用户自主解决问题，而非直接生成代码片段。该模式特别适合需要深入理解框架或编程语言的学习者、教育工作者及开发者。

两项创新功能现已集成于Colab的Gemini对话界面，用户可通过聊天窗口快速启用。谷歌表示，此次升级致力于让AI助手更灵活地适配教学、协作与个人学习场景，推动知识共享与技能成长。

中文翻译：

介绍学习模式：您在Google Colab中的个人编程导师
Google Colab正在扩展其Gemini集成功能，推出两项全新改进，帮助您在Colab中自定义Gemini智能助手。通过自定义指令和学习模式，您可以更灵活地掌控Colab中Gemini的工作方式，让它更好地助力您的学习。

自定义指令存储在笔记本层级，允许笔记本作者定制专属的Colab Gemini AI助手，并将个性化助手分享给Colab社区。学习模式则将Gemini转变为个人编程导师，提供循序渐进的指导，而非直接为您编写代码。

无论您是温习新框架的资深开发者、设计课程的教育者，还是初涉编程的学生，自定义指令和学习模式都能让Gemini成为更灵活、独一无二的助手。

定制您的偏好
通过自定义指令，您可以根据个人工作流程、课程或项目需求添加上下文信息。
您有偏好的代码风格吗？希望助手在提供建议时参考课程大纲吗？需要Gemini教学时始终使用特定代码库吗？将这些偏好设为自定义指令即可。您可以直接在Gemini聊天框内切换设置，并将其保存至笔记本中。

在学习模式中提升技能
是否曾希望Gemini引导您思考，而非直接给出答案？
学习模式基于自定义指令实现，它不会用大段可复制的代码直接解决问题，而是通过分步指导来解答编程疑问：拆解复杂主题、阐释核心概念，帮助您真正掌握技能。

您可以在Colab的Gemini聊天窗口中开启学习模式。对于希望深入理解新框架或编程语言的学习者而言，这将是完美的伙伴。

下方提供了一些示例笔记本。您可以在其中与Gemini Colab助手对话（这些笔记本已预设启用学习模式），它将引导您完成Python编程练习。

分享您的智慧
Colab的魅力之一在于其社区生态和作品共享能力，我们在设计这些功能时也充分考虑了分享需求。

自定义指令直接存储在您的笔记本中，并适用于该笔记本未来的所有Gemini对话。当您将笔记本分享给同事或更广泛的社区时，这些指令会随之传递。这确保了任何协作者或学习者都能体验到您精心设计的定制化AI辅助。

我们期待看到您如何运用自定义指令和学习模式去探索、创造并分享您的成果。

英文来源：

Introducing Learn Mode: your personal coding tutor in Google Colab
Google Colab is expanding its Gemini integration with two new improvements that will help you customize your Gemini agent in Colab. Custom Instructions and Learn Mode give you more control over how Gemini in Colab works and how it helps you learn.
Custom Instructions, stored at the notebook level, allow notebook authors to tailor their Colab Gemini AI assistant and share their personalized assistants with the Colab community. Learn Mode turns Gemini into a personal coding tutor that provides step-by-step guidance instead of writing code for you.
Whether you're a seasoned developer brushing up on a new framework, an educator designing coursework or a student who’s new to coding, Custom Instructions and Learn Mode make Gemini a more flexible, one-of-a-kind assistant.
Customize your preferences
Add context to fit your personal workflow, class or project needs with Custom Instructions.
Do you have a preferred coding style? Do you want your agent to be aware of your class syllabus when making recommendations? Do you want Gemini to always use a specific library when teaching you? Add those preferences as Custom Instructions. Toggle them directly from the Gemini chat box and save them to your notebook.
Build your skills in Learn Mode
Have you ever wanted Gemini to guide you rather than just give you the answer?
Instead of solving your problem with a block of copy-paste code, Learn Mode — which is powered by Custom Instructions — answers coding questions with step-by-step instructions that break down complex topics, explain the underlying concepts and help you develop your skills.
You can turn on Learn Mode from the Gemini chat window in Colab. It's the perfect companion for anyone looking to deepen their understanding of new frameworks or programming languages.
You can see some example notebooks below. Start a chat with the Gemini Colab agent, which is preprogrammed to use Learn Mode in these notebooks, and it will guide you through the Python programming exercises.
Share your knowledge
One of the best parts about Colab is the community and the ability to share your work, and we built these features with sharing in mind.
Custom Instructions are stored directly within your notebook and applied to future Gemini chats in that notebook. When you share your notebook with colleagues or the broader community, those instructions travel with it. This ensures that anyone collaborating on or learning from your notebook gets the tailored AI experience you designed.
We can't wait to see how you use Custom Instructions and Learn Mode to discover, build and share your work.