Sreehari suresh

( ͡° ͜ʖ ͡°)

wannabe design engineer / idk

[Sreehari suresh @raspberry__1]

Whatsss damn

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[Sreehari suresh @raspberry__1]

lmao durandhar irl?

[ABC News @ABC]

NEW: Eileen Wang, the mayor of Arcadia, California, has been charged with acting as an illegal foreign agent for China, the Justice Department announced on Monday.

Wang agreed to plead guilty, the Justice Department said. https://t.co/61VZ41RJMF

[Sreehari suresh @raspberry__1]

how cool is this.

[Brian Chesky @bchesky]

@benhylak Let’s talk

[Sreehari suresh @raspberry__1]

I’m really bad with names. I forgot some friends’ names and tried to play it off, but they just wouldn’t stop bugging me, asking me to tell their name. Eventually, I had to admit I had forgotten. It was the same with a teacher.😭

Embarrassing

[Ekaeo @Ekaeoq]

It's this guy and it's not even close:

[Sreehari suresh @raspberry__1]

you go to school on a rainy day, the classrooms are soo dark because of the rain , there is no power. Classes haven’t begun. You are just there enjoying the rain and wind. An hour later you hear the state has declared leave and you head home with the biggest smile in your face

[Sreehari suresh @raspberry__1]

the site is finally coming together.
with a few easter eggs here and there 😉
https://t.co/yHYnd6qOXC

[Sreehari suresh @raspberry__1]

@bryan_johnson JESUS..

Im not even christian

[Sreehari suresh @raspberry__1]

lmao..I kinda built one myself last year for the same exact reason using obsidian as aplugin, its a simple rag system but it worked nicely.

it also had an inbuilt chat ui as well.

and then I abandoned it. Classic me

[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.

[Sreehari suresh @raspberry__1]

Wait what?

[will depue @willdepue]

shoutout bihar

[pH @pHequals7]

2026: how i learnt my masala dosa depends on the strait of hormuz

[Sreehari suresh @raspberry__1]

@levelsio Wtf

[Sreehari suresh @raspberry__1]

@IRCTCofficial @RailMinIndia @GMSRailway

[Sreehari suresh @raspberry__1]

What the hell @IRCTCofficial my ticket was showing rac 11 in the morning, and all of a sudden it's now waiting list? And the ticket got unconfirmed.

My whole trip is ruined.

[Sreehari suresh @raspberry__1]

@CinemasBroadway dream

[Sreehari suresh @raspberry__1]

@poettoearth Vv clean

[Sreehari suresh @raspberry__1]

@StreetFashion01 feels like a wes anderson movie

[Sreehari suresh @raspberry__1]

bro antigravity is such a terrible Vibe code app. It's an ass code app. Bro be rage Baiting me all the time. Wtf

[Sreehari suresh @raspberry__1]

@kilimeeeen @w0rdgenerator this is what it feels like when you play ping pong.

[Sreehari suresh @raspberry__1]

@Aurelien_Gz damn it even has the smudges from the clicks Imao.