Martin Henry (@emailnitram) header
Martin Henry (@emailnitram) profile picture

Bay area

Joined July 2011
Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)Photo by Martin Henry (@emailnitram)
Martin Henry retweeted
Wow, this tweet went very viral! I wanted share a possibly slightly improved version of the tweet in an "idea file". The idea of the idea file is that in this era of LLM agents, there is less of a point/need of sharing the specific code/app, you just share the idea, then the other person's agent customizes & builds it for your specific needs. So here's the idea in a gist format: gist.github.com/karpathy/442… You can give this to your agent and it can build you your own LLM wiki and guide you on how to use it etc. It's intentionally kept a little bit abstract/vague because there are so many directions to take this in. And ofc, people can adjust the idea or contribute their own in the Discussion which is cool.
LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.
1,121
2,824
685
26,776
7,352,747
Martin Henry retweeted
/buddy
466
164
188
4,008
565,708
💡 React hooks tip: how to properly use useCallback.
6
75
406
⚛️ React useEffect cheat sheet
22
156
7
822
Martin Henry retweeted
Name your JavaScript exports! This ensures they'll show up named in DevTools and in compile time errors.
26
306
13
1,563
Martin Henry retweeted
Ever tried debugging an element that keeps disappearing when it loses focus once you start using devtools? Well dang me to heck there is an "Emulate a focused page" option in @ChromeDevTools just for that 🤯 (you can get it from the [⌘]+[P] Command Menu, or Global Preferences)
57
1,056
109
3,783
CSS transitions
7
211
7
1,077
z-index & stacking contexts
7
230
11
1,085
Studying to become #AWSCertified? I got you covered: 🚀 free AWS Certified Cloud Practitioner youtu.be/3hLmDS179YE 🚀 free AWS Solutions Architect Associate youtu.be/Ia-UEYYR44s 🚀 free AWS Developer Associate youtu.be/RrKRN9zRBWs
16
152
11
456
Martin Henry retweeted
JavaScript tip: did you know you can listen for a *single occurrence* of an event using addEventListener? You can specify a configuration object as a third parameter.
15
261
8
1,042
Martin Henry retweeted
Fixing memory leaks in web applications: https://teapawt.net/t.co/gtn6DKu1cx ~ a guide to common causes of JavaScript memory leaks, how to identify and fix them 🌊🔧
8
478
14
1,496
Martin Henry retweeted
JavaScript loading priorities in Chrome: https://teapawt.net/t.co/O3Qg2eYROJ ~ how do <script>, <script defer>, <link rel=preload> + <script async> & friends load then execute? Useful table.
12
767
15
2,265
For Valentine's Day this year, we're asking you to help us give back. $15 would keep several people dry. $30 would buy water for a week of outreach. $110 gives socks to 120 people. Please Venmo @ KtownforAll or you can also go shopping at tinyurl.com/ktownwishlist Thank you 💜❤️
1
27
12
32
Martin Henry retweeted
Browsers can animate at least 4 things cheaply: position, scale, rotation & opacity. csstriggers.com/ is still useful for checking if CSS triggers re-layouts or reflows.
7
269
2
1,150
Awesome summary of modules! If you thought my cheatsheet was useful, you won’t want to miss this short clip 👏👏👏
Take 104 seconds learn about #JavaScript (ES) modules and grab this awesome cheatsheet from @samantha_ming #100SecondsOfCode 🎞️youtu.be/qgRUr-YUk1Q
6
59
1
293
Are you using @Cypress_io? I have a present for you: github.com/testing-library/c… This will make your tests WAY easier to read and write and a lot more resilient. Bonus: your Cypress E2E tests will look a lot like your Jest integration/unit tests if you use @TestingLib there too!
22
170
5
654
Martin Henry retweeted
Git tip I wish I'd discovered ten years ago: if you `git config --global diff.noprefix true` it removes the silly `a/` and `b/` prefixes so that when you double-click select one to copy, you get a usable filename instead of a mangled path.
29
607
49
1,767
Martin Henry retweeted
When do which shell startup files get executed? blog.flowblok.id.au/2013-02/…
1
22
74
every commit has a parent
9
67
3
382
more git fundamentals for @ohshitgit: a branch is a pointer to a commit
17
580
27
2,139