๐ค Summary generated by GPT.
miniWorldModel is a personal, long-term knowledge base built in Obsidian. It began as a set of notes for undergraduate computer science courses and gradually evolved into a structured attempt to organize my understanding of the world.
The notes are organized as a single, large mind map: folders represent higher-level abstractions, individual Markdown files act as nodes, and dense bidirectional linksโboth in text and in diagramsโform the connections between ideas. Rather than aiming for depth in every topic, the project focuses on unifying knowledge (especially computer science) into a coherent, integrated framework.
This repository does not claim to describe the universe as it objectively is. Instead, it reflects how I currently understand it. In that sense, miniWorldModel is a snapshot of my internal world modelโan evolving representation of reality shaped by learning, reflection, and revision.
While learning computer science, I wanted a way to take notes so that my thoughts sorted out. I tried several notebook applications and mind-mapping tools, but they felt weak and restrictive for the kind of structured thinking I needed. Thanks to Obsidianโs powerful customization and linking capabilities, I decided to build my own note system on it. This repository is the result.
The notes are structured as a giant mind map. Each md file acts as a leaf node, while folders represent parent nodes in the hierarchy. Inside each md file, there are many ๐ bidirectional internal links at the paragraph level, which function like the connections between nodes in a traditional mind map.
Below is an example of this organizational structure. It does not reflect the actual contents of my notebook, but it demonstrates the same organizational principles. In the real one there are many, many more such internal links.
(Generated by GPT-5.2)
WORLD MODEL (miniWorldModel)
โ
โโโ Mind & Cognition
โ โโโ Perception
โ โ โโโ Senses
โ โ โโโ Representation
โ โโโ Consciousness
โ โ โโโ Awareness
โ โ โโโ Qualia
โ โ โโโ Self-Model
โ โโโ Intelligence
โ โ โโโ Learning ---------------------
โ โ โโโ Reasoning |
โ โ โโโ Prediction |
โ โโโ Emotions & Motivation | This is an
โ | internal link
โโโ Knowledge | between notes
โ โโโ Epistemology |
โ โ โโโ Truth |
โ โ โโโ Belief <--------------------
โ โ โโโ Uncertainty
โ โ โโโ Bias
โ โโโ Science
โ โ โโโ Method
โ โ โโโ Models
โ โโโ Mathematics
โ โ โโโ Logic
โ โ โโโ Structures
โ โ โโโ Abstractions
โ โโโ Philosophy
โ
โโโ Meta
โโโ Models of Models
โโโ Assumptions
โโโ Unknowns
โโโ Open Questions
โโโ Revision RulesInitially, this notebook was intended only as a place to store notes for my undergraduate computer science courses. As it grew larger and my learning expanded, it gradually evolved into this giant knowledge base that tries to framework all the knowledge of the universe (although the notebook is still largely about computer science). I do not claim that these notes accurately represent the universe as it truly is. Rather, they represent the way I understand the universe. I write them just to sort out my own thoughts.
For this reason, I named this notebook miniWorldModel. It is โminiโ because it is incomplete and approximate, and a โworld modelโ because it reflects my internal representation of the actual universe. This internal thinking is what defines me as me, as in Renรฉ Descartesโ famous statement: โI think, therefore I amโ โ or ๐ cogito, ergo sum in Latin. In other words, this notebook is a record of how I think about the universe; and in that sense, it is a record of my existence within it. If you are reading this, you are reading me.
Tip
I have thought extensively about self-awareness and intelligence, which you can see reflected in my notes on โ Universe, Self-Awareness, and Intelligence. Below is a diagram illustrating the relationship between a world model and intelligence.
A diagram illustrates what is world model and what is thought as the meaning of "intelligence".
Murphy, K. (2025). Reinforcement Learning: An Overview (No. arXiv:2412.05265). arXiv.
https://doi.org/10.48550/arXiv.2412.05265
This notebook is not meant to be deep in every topic. Instead, it aims to cover as many domains as possible and, more importantly, to organize them under a single cognitive frameworkโmy own cognitive model. That said, some areas do have a certain level of depth, particularly in some directions under computer science, as I am still actively learning and refining my understanding.
This repository has been public since its creation three years ago. However, the notes themselves have existed locally on my computer for more than four years. Only recently did I decide to make the repository more presentable, as it could genuinely be useful to others. That is why I am writing this introduction now. I hesitate to say that I am โreleasingโ it, since it has already existed publicly in some form for years.
I was born and raised in ๐ Xinjiang, China, so you may occasionally encounter Chinese contents among notes, sometimes mixed with English. However, most of the contents are written in English, making it accessible to most readers. (ไธ่ฟ็ธๅฏนๆฅ่ฏด่ฟๆฏๆๅพๅคๅ ณไบไธญๅฝ็ๆ่ๅๆไฝไธบไธไธชไธญๅฝไบบ็่ง่ง็ๅไบใๆฏ็ซๆๆฏไธญๅฝไบบใๆ็ฑๆ่ชๅทฑ๏ผๆไปฅๆ็ฑๆ็ๅฝๅฎถใ) That being said, the note will keep English as the only primary language to use.
In the end, miniWorldModel is simply my attempt to organize what I have learned and the thoughts I have had along the way. Although I strive to be precise and rigorous, there are inevitably errors and inaccuracies.
To whoever reading this, I hope my notes can serve as an inspiration.
All notes in this repository are written entirely in ๐ Obsidian. As a result, they are best viewed within Obsidian, where the structure, links, and diagrams appear exactly as I intended (for reference, I am using the ๐ Atom theme). When viewed directly on the GitHub webpage, some content may appear poorly formatted or even unreadable. For the best experience, I strongly recommend cloning the repository locally and opening it in Obsidian.
An essential plugin for viewing these notes is ๐ Excalidraw, as many diagrams in this repository are created using it. Without this plugin, you will not be able to see those diagrams, which would mean missing a significant part of the content.
Many Excalidraw diagrams also contain internal links, allowing you to jump directly from one diagram to another diagram or note file. This makes diagrams an integral part of the navigation, not just visual aids.
Another important note: this repository is, at its core, a personal knowledge base. For some reasons, some files are intentionally not shared publicly, which may occasionally lead to broken links or missing context when browsing related notes. Unfortunately, I cannot fully resolve this issue. However, unless the content is restricted due to personal privacy or copyright concerns, there is usually an external link provided to the original resource, so you can still access the intended material.
These notes have been created continuously since my first semester at university. As a result, some early notesโespecially those I rarely revisitโmay be of lower quality.
Final tip (and this is how I read my own notes): ==always understand the structure before diving into the details==. When opening a file, first look at its folder hierarchyโits parent folder, sibling folders, and higher-level directories. Similarly, when reading a note, start with its outline and use it as a guide to navigate the content.
For those notes in a mature quality, they follow the following template in general:
# Title
## Res
### Related Topics
### Other Resources
## Intro
## RefRes lists related resources (internal notes or external webpages, books, etc.) to the contents of this file, while Ref lists the references in the file. Intro is reserved for the body of the file. Usually, between Res and Ref there would be much more than just an Intro.
Specifically, under Res, Related Topics lists internal notes in this notebook. For the books, videos, websites, etc, they are in the rest sub-sections of Res (for example, Other Resources).
This notebook is large. Although it is well-organized as a mind map with all the internal links, it could still be hard to quickly navigate around them. This is why we need to search.
During my writing, I mainly use three searching methods:
- file search: in obsidian,
ctl-oopens the native file search bar. - text search: in obsidian, I use a plugin ๐ OmniSearch. This allows user to search at text level in all files.
- tag search: in my notes, sometimes I will use tags for contents that's a FAQ, or those paragraphs /sub-sections where I compares two concepts. For example, if you are in obsidian, try open the search bar (
cmd-shift-f) and searchtag: #vim. Just note that tag search might not give you everything you want to search because I don't write tags everywhere.
In addition to my favorite searching methods, there are many other search options natively supported by obsidian.
When I first started taking these notes (around 2021), one major motivation was how difficult it was for me to even find the right tools and resources for starting to learn computer science. Especially, if you were in China, the internet is banned (โ Proxy Technology & GFW ๐จ๐ณ). In the beginning, this notebook functioned largely as a curated list (like an โ ๐คฏ Awesome List) โcollecting useful materials and references for topics I encountered along the way, and organizing them in a connected fashion. Today, in the era of ChatGPT, that gap is gone forever (๐ the bitter lesson!).
This naturally raises a question: what is the value of a notebook like this now, when we can simply ask an AI whenever we do not know something? Going further, it may even seem that we are approaching a future where knowing things ourselves (into the trivial implementing details) is no longer necessary, because AI can handle them for us.
The answer is the same one hinted at earlier. miniWorldModel is not about being deep in every individual topic. Instead, it aims to unify (my personal) knowledgeโespecially undergraduate-level computer scienceโinto a single, integrated framework. It provides a systematic, high-level view of how everything fits together. For me personally, it is also the primary way I organize my own thinking: not just storing information, but structuring understanding.
Lastly, again hinted at earlier, this notebook is me. It's my thoughts of the universe. It's the record of my life. Asking "what is the meaning of this notebook" is equivalent to "what is the meaning of my existence in this universe". And that is for sure AI cannot replace.
Like said earlier, miniWorldModel is a well-organized notebook that acts like a giant mind map. Hence, to start reading the notebook one needs simply start with the only root file (the root node of the mind map) as everything unfolds here: โ Universe, Self-Awareness, and Intelligence. In this root file, I summarize my cognitive model of the universe as a unified framework (thus, the whole structure of this notebook).
That being said, I give a very brief (and incomplete) index to miniWorldModel below. You can view them as the entry points to the miniWorldModel.
- If you follow my suggestions in previous sections and explore notes hierarchy and links from these entry points, eventually all thinking comes back to the root file, where we discuss Universe, self-awareness, and intelligence, i.e. the being of ourselves.
Tip
Bare in mind this whole notebook is still heavily about computer science.
Also, these notes are simply a reflection of my personal thinking all along the time. Many of them are incomplete, and inevitably contain errors and inaccuracy. Please be mercy with me. โ๏ธ
โ Universe, Self-Awareness, and Intelligence
- โ Philosophy & Its History
- โ Human History
- โ Arts & Humanities
- โ Language & Literature
- โ Music
- โ Mathematics
- โ Science & Application
- โ Natural Science & Engineering and Technology
- Information Science & Computer Science and Engineering
- Computer Science Overview
- Computer Science Core Knowledge
- Computer Science Specific Directions
- โ Artificial Intelligence
- โ Software Engineering
- โ CyberSecurity
- Opportunities & Career Development
- Information Science & Computer Science and Engineering
- โ Social Science
- โ Natural Science & Engineering and Technology
As I repeatedly stress, this entire notebook is structured as a mind map, on top of which are enormous links connecting those leaf nodes, like illustrated in the beginning of this readme. Although the entire hierarchy and all the relationships between files are clear in my head, I find it very hard to present to others of such structure. One big reason is the lack of existing tools /software to visualize the whole notebook.
Below are the screenshots (if you are inside obsidian you can scroll right to see more pics, since they are displayed in carousel style) from the native plugin ๐ graph view of the notebook. This graph view looks pretty cool, but it doesn't show the nodes in the hierarchical way as they are.
Below are screenshots showing the structure of folders as a mind map (no files, just folders). It was generated from GPT-generated program.
By my intention, a really good visualization would be the combination of the above two: hierarchy + linking. Unfortunately, right now I cannot show this to you directly.
At the moment miniWorldModel is only in its beta version. I don't expect too much attention in the short future as well, since my studying life is still pretty hectic and I may not have time for all the promotion and stuff. But anyway I made some starts in getting it ready for others to read. Hopefully this note would look nicer in the future.
Again, the notes might contain lots of errors and inaccuracy. Please read it as a friend sharing his study notes to you.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
