Tyler Neylon's Homepage
This is an experiment to see if Claude can write poetry. I collaborated with Claude Opus 4.5 to create a structure called a corona, which consists of 15 sonnets, the final sonnet being composed of the first lines of fourteen preceding pieces.
A simple model of how digital minds might work.
This is a 3d puzzle you can build with Legos, and might be able to solve in your head.
A thoughtful book summary and review for Cal Newport's So Good They Can't Ignore You.
While working on a prototype machine learning system in Python, I hit a concurrency deadlock that seemed at first to be impossible. This is the story of that bug.
This short article describes a chalkboard method for multiplying (that is, composing) permutations written in cycle notation.
This is a collection of notes about permutations — all the different ways of ordering a set of distinct elements.
This article is an illustrated tour of neural networks in their primordial, untrained state. Neural networks are notoriously difficult beasts to understand. My aim is to provide a peek into the inherent beauty of this challenging world, and to build your intuition for how to set up neural networks through informed hyperparameter choices.
Locality-sensitive hashing (LSH) is a set of techniques that dramatically speed up search-for-neighbors or near-duplication detection on data. These techniques can be used, for example, to filter out duplicates of scraped web pages at an impressive speed, or to perform near-constant-time lookups of nearby points from a geospatial data set.
What does it mean for something to be true? What can we do to verify that something is true? How can we discover new truths? This article is about a conceptual model of truth that may help answer these questions.