\(\mu\)P as Optimal Transport in a Vanilla MLP
Deriving \(\mu\)P as the unique scaling maximizing Wasserstein transport under stability constraints
Posts
A collection of thoughts, notes, and small projects, updated infrequently. This page is mostly kept updated for my own reference (mostly to keep track of my interests, which change pretty frequently), but hopefully some of the posts are interesting to others.
Jan 05, 2026
Dec 12, 2025
Room Auctions
Designing an auction system for rooms in our house
Nov 16, 2025
Deep Learning Training Dynamics are not Fisher Geodesics
A cool bit of math that falls apart empirically
Nov 02, 2025
A Category Theoretic View of Machine Learning
Using symmetries to better understand how to design attention kernels
Sep 13, 2025
Every Model is Kernel Ridge Regression
Some notes on Domingos' paper: Every Model Learned by Gradient Descent Is Approximately a Kernel Machine
Aug 23, 2025
A Markov Picture of Heat
A derivation of the heat equation from a stochastic model of energy exchange
Jun 11, 2025
Eigenlearning
Some notes on the Eigenlearning paper by Simon et al.
Jun 03, 2025
Reproducing Kernel Hilbert Spaces
Building up RKHSs from the Riesz Representation Theorem
Apr 19, 2025
Cookie Magic
How I got ~$500 worth of cookies for free
Apr 11, 2025
Neural Tangent Kernels
A rough derivation of the neural tangent kernel
Jan 19, 2025
Hopfield Networks
A quick implementation of a Hopfield network in Python