Dmitry Balabka dbalabka

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

Create a file in C:\bin\toast.ps1 from source https://gist.github.com/dend/5ae8a70678e3a35d02ecd39c12f99110:
```
curl https://gist.githubusercontent.com/dend/5ae8a70678e3a35d02ecd39c12f99110/raw -o /mnt/c/bin/toast.ps1
```

Create ~/.local/bin/notify-send from the following source:

powershell.exe "& { . C:\\bin\\toast.ps1; Show-Notification -ToastTitle \"$1\" -ToastText \"$2\" }"

like this:

Anti-hype LLM reading list

Goals: Add links that are reasonable and good explanations of how stuff works. No hype and no vendor content if possible. Practical first-hand accounts of models in prod eagerly sought.

Foundational Concepts

Pre-Transformer Models

`<script> async, defer, async defer, module, nomodule, src, inline` - the cheat sheet

This document is a comparison of various ways the <script> tags in HTML are processed depending on the attributes set.

If you ever wondered when to use inline <script async type="module"> and when <script nomodule defer src="...">, you're in the good place!

Note that this article is about <script>s inserted in the HTML; the behavior of <script>s inserted at runtime is slightly different - see Deep dive into the murky waters of script loading by Jake Archibald (2013)

Dependency Injection in Functional Programming

There exist several DI frameworks / libraries in the Scala ecosystem. But the more functional code you write the more you'll realize there's no need to use any of them.

A few of the most claimed benefits are the following:

Dependency Injection.
Life cycle management.
Dependency graph rewriting.

High-Performance Matrix Multiplication

This is a short post that explains how to write a high-performance matrix multiplication program on modern processors. In this tutorial I will use a single core of the Skylake-client CPU with AVX2, but the principles in this post also apply to other processors with different instruction sets (such as AVX512).

Intro

Matrix multiplication is a mathematical operation that defines the product of

Google Public NTP [AS15169]:

time.google.com

time1.google.com

time2.google.com

time3.google.com

	function Show-Notification {
	[cmdletbinding()]
	Param (
	[string]
	$ToastTitle,
	[string]
	[parameter(ValueFromPipeline)]
	$ToastText
	)

	# Presto failed with error
	# 82) Error injecting constructor, java.lang.IllegalStateException: INVARIANT: current node not returned from service selector


	coordinator=true
	node-scheduler.include-coordinator=true
	http-server.http.port=8080
	query.max-memory=5GB
	query.max-memory-per-node=1GB
	query.max-total-memory-per-node=2GB