Yaroslav Petrashko

20 // grad student @columbia // prev ML @ornl // 4x hackathon winner // orthodox

[Yaroslav Petrashko @YaroPetrashko]

saturday evening study sesh

implementing 1 ML algorithm/concept from scratch per week challenge begins today :)

inspired by @saurabhalonee

[Yaroslav Petrashko @YaroPetrashko]

Quick life update: Moving to NYC soon to start my Master’s at @Columbia

Very excited!

[Yaroslav Petrashko @YaroPetrashko]

.@BillJiao930 the 🐐

warmest congratulations on the launch!

[Y Combinator @ycombinator]

General Instinct (@gen_instinct) deploys frontier AI models onto constrained edge hardware, helping robotics and physical AI teams run low-latency models offline on devices like Jetsons, mobile NPUs, and ARM CPUs.

Congrats on the launch, @BillJiao930 & @Guanming717!

https://t.co/JKEpLrSTdj

[Yaroslav Petrashko @YaroPetrashko]

@neuralkian fire, hope to read it throughout the next couple weeks.

thanks for your work!

[Yaroslav Petrashko @YaroPetrashko]

time to come back to this banger

Agentic Zettelkasten is going to be the industry standard very soon

jump on the bandwagon early while you still can👀

[Andrej Karpathy @karpathy]

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.

[Yaroslav Petrashko @YaroPetrashko]

@karpathy @liminal_warmth How about research papers? What's the best format to use when feeding them to the wiki?

[Yaroslav Petrashko @YaroPetrashko]

@edwardjxli @hackwithtrees cracked

[Yaroslav Petrashko @YaroPetrashko]

TreeHacks 2026 was lit🌲

Happy to have met hundreds of cracked builders🙌

[Yaroslav Petrashko @YaroPetrashko]

@thorchh @AnthropicAI @hackwithtrees lets go thor! ⚒️

[Yaroslav Petrashko @YaroPetrashko]

@qtzx06 @hackwithtrees W

[Kye Gomez (swarms) @KyeGomezB]

Treehacks 2026

Met so many incredibly intelligent builders today at Stanford, many of them building unique applications with Agents.

If you were at Treehacks and we didn’t get to meet reach out, we’re recruiting agent engineers to work on our various protocols and framework

With @YaroPetrashko

Godspeed

[Yaroslav Petrashko @YaroPetrashko]

Met the goat @gabriel1 at TreeHacks🇸🇪🇺🇦

[Yaroslav Petrashko @YaroPetrashko]

@stevekrouse And they named mount Everest after his great-great-granduncle George Everest (which is in his middle name as well) 👀

[Yaroslav Petrashko @YaroPetrashko]

@kaarelkaarelson @zfellows

[Yaroslav Petrashko @YaroPetrashko]

With @DavidSHolz at @hackwithtrees 2026🤝

[Yaroslav Petrashko @YaroPetrashko]

@agniv_s @hackwithtrees @Kalshi @Polymarket @ManifoldMarkets very sick👀🔥

[Yaroslav Petrashko @YaroPetrashko]

@cory great conversations, amazing people, thanks a lot!🙌

[Yaroslav Petrashko @YaroPetrashko]

crazy time we live in
@sama

[Yaroslav Petrashko @YaroPetrashko]

@cdngdev cook incoming

[Yaroslav Petrashko @YaroPetrashko]

@cory will be competing this weekend and would absolutely love to join!

[Yaroslav Petrashko @YaroPetrashko]

hyped

[TreeHacks @hackwithtrees]

Announcing our TreeHacks 2026 Keynote Speaker: Sam Altman
@sama

February 13 - 15, 2026.