Stories by jxmorris12

Kullback-Leibler (KL) Is All You Need

How Do Language Models Put Attention Weights over Long Context?

Chess Engines: A Zero to One

A Personal History of Legion, by Way of Its Papers

Bananagrams Is NP-Complete

A Better Lesson

Seeking the Productive Life: Some Details of My Personal Infrastructure

A Neighborhood with Friends

How to graduate your PhD when you have no hope

PyTorch Word Embeddings Tutorial

Integer Tokenization Is Insane (2023)

Diffusion models from scratch, from a new theoretical perspective

A Refined Similarity-Based Bigram Model

Speculative Sampling

An Introduction to Optimization: Combinatorial Optimization

Definite Optimism as Human Capital

Singular Value Decomposition Part 1: Perspectives on Linear Algebra

Singular Value Decomposition as Simply as Possible

Too much efficiency makes everything worse: overfitting and Goodhart's law

Detecting Mismatches in Machine-Learning Systems

12
13
14
15
16
17