karpathy

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.

hrbrmstr

viridis_pal <- function(alpha=1) {
  function(n) {
    viridis(n, alpha)
  }
}


scale_color_viridis <- function(..., alpha=1, discrete=TRUE) {
  if (discrete) {
    discrete_scale("colour", "viridis", viridis_pal(alpha), ...)

jennybc

#' ---
#' output:
#'   html_document:
#'     keep_md: TRUE
#' ---

#+ include = FALSE
library(dplyr)

#' Responses to [my

meyera

## Curated stuff:
## * MetaCran: http://www.r-pkg.org/
## * List of Packages gathered by Garret G.: https://github.com/rstudio/RStartHere
## * List of popular packages https://awesome-r.com/
## * List of DataScience R tutorials https://github.com/ujjwalkarn/DataScienceR
## * List of machine elearning tutorials by subject: https://ujjwalkarn.github.io/Machine-Learning-Tutorials/

## Reproducible package management for R:
install.packages("devtools")
install.packages("tidyverse")

araastat

#’ Create a Kaplan-Meier plot using ggplot2
#’
#’ @param sfit a \code{\link[survival]{survfit}} object
#’ @param returns logical: if \code{TRUE}, return an ggplot object
#’ @param xlabs x-axis label
#’ @param ylabs y-axis label
#’ @param ystratalabs The strata labels. \code{Default = levels(summary(sfit)$strata)}
#’ @param ystrataname The legend name. Default = “Strata”
#’ @param timeby numeric: control the granularity along the time-axis
#’ @param main plot title