yoavg

Learning LLMs in 2025

So you know how the transformer works, and you know basic ML/DL, and you want to learn more about LLMs. One way to go is looking into the various "algorithmic" stuff (optimization algorithms, RL, DPO, etc). Lot's of materials on that. But the interesting stuff is (in my opinion at least) not there.

This is an attempt to collect a list of academic (or academic-like) materials that explore LLMs from other directions, and focus on the non-ML-algorithmic aspects.

Courses

• David Chiang's Theory of Neural Networks course.
• This is not primarily LLMs, but does have substantial section on Transformers. Formal/Theory. More of a book than a course.

marcosci

library(LaCroixColoR)
library(sf)
library(fasterize)
library(rayshader)
library(raster)
library(exactextractr)
library(rayrender)

# load raster into R
elevation <- raster("Olympus_Mons_ortho-image.tif")

walkerjeffd

library(tidyverse)
library(dataRetrieval)
library(rayshader)

# install {av} to render movie
# install.packages("av")

# fetch daily streamflow from USGS NWIS
df_nwis <- readNWISdv(siteNumbers = "01059000", parameterCd = "00060") %>% 
  renameNWISColumns() %>% 

medelman17

import * as fs from 'fs';
import * as path from 'path';
import * as cdk from '@aws-cdk/core';
import * as lambda from '@aws-cdk/aws-lambda-nodejs';
import * as kms from '@aws-cdk/aws-kms';
import * as secretsmanager from '@aws-cdk/aws-secretsmanager';
import * as acm from '@aws-cdk/aws-certificatemanager';
import * as alias from '@aws-cdk/aws-route53-targets';
import * as iam from '@aws-cdk/aws-iam';
import * as api from '@aws-cdk/aws-apigatewayv2';

tylermorganwall

library(deSolve)
library(rayrender)
parameters = c(a = 0.2, b = 4, c = 8, d = 1)
state = c(X = 1, Y = 0, Z = 1)

Lorenz = function(t, state, parameters) {
  with(as.list(c(state, parameters)), {
    dX = -Y^2 - Z^2 - a*X + a*c
    dY = X * Y - b * X * Z - Y + d
    dZ = b * X * Y + X * Z - Z

tylermorganwall

library(raster)
library(rayshader)
#Load QGIS georeference image (see https://www.qgistutorials.com/en/docs/3/georeferencing_basics.html)
testindia = raster::stack("1870_southern-india_modified.tif")

#Set bounding box for final map (cut off edges without data, introduced via reprojection)
india_bb = raster::extent(c(68,92,1,20))
cropped_india = raster::crop(testindia, india_bb)

#Convert to RGB array

whitequark

#lang rosette/safe
(require rosette/lib/angelic
         (only-in racket/base syntax->datum))

(define-syntax-rule (define-distinct v x ...)
 (begin
  (define x v) ...
  (assert (distinct? x ...))))
(define-syntax-rule (list-choices x ... sol)
  (list (list (evaluate x sol) (syntax->datum #'x)) ...))

cowboy

/*
 * Abstraction.js 2020 Edition
 *
 * Copyright (c) 2020 "Cowboy" Ben Alman
 * Licensed under the MIT license.
 * http://benalman.com/about/license/
 */

let $elseif, $else, $if = state => state ? (
  $elseif = () => () => {},

andrewroberts

// dev: andrewroberts.net

// Replace this with ID of your template document.
var TEMPLATE_ID = ''

// var TEMPLATE_ID = '1wtGEp27HNEVwImeh2as7bRNw-tO4HkwPGcAsTrSNTPc' // Demo template
// Demo script - http://bit.ly/createPDF
 
// You can specify a name for the new PDF file here, or leave empty to use the 
// name of the template. Placeholders can also be placed in here.

samueleresca

import {Component} from '@nestjs/common';
import {InjectRepository} from '@nestjs/typeorm';
import {ILabelService} from './ILabelService';
import {Repository} from 'typeorm';
import {Label} from '../Models/Label';


@Component()
export class LabelsService implements ILabelService {