Notes from the frontier.

BlogAug 2026

The Weight Comes Last

The leading terms cancel modulo q, and the degree left behind is settled by whether an addition in base q carries — the reduction Evan Chen and Ken Ono use to prove Thakur's three hypotheses, open since 2009.

BlogJul 2026

The Address Before the Room

One family is a third the size of the other — but that third is a doorway, not the matching that has to be built behind it.

BlogJul 2026

The Same Ground

In the right coordinate, Kaprekar's process is not a different puzzle in every base — it is one operation: doubling.

BlogJul 2026

The Figure and the Remainder

A multiplicity that reads as the blur of cancellation comes into focus, in natural families, as an exact count of bounded Dyck-path structures.

BlogJun 2026

The Reveal

In the hardest window for lattice triangles, a density-1 result rules out almost every candidate.

BlogMay 2026

Infinitely Rare

When can Ramanujan's tau function take a prime value? Almost never.

BlogMar 2026

Axplorer

Democratizing the search for mathematical constructions.

BlogMar 2026

Releasing AXLE

Our Infrastructure for Mathematical Reasoning at Scale.

BlogFeb 2026

Proof of Concept

An autonomous, verified proof on an open conjecture linking syzygies to Ramanujan’s identities

BlogFeb 2026

Primes: Almost surely, partially, regular

A 350-year-old question, a new theorem, and a proof written by a machine

BlogJan 2026

From Seeing Why to Checking Everything

AxiomProver Solves All Problems at Putnam 2025: Proof Release & Commentary

BlogDec 2025

Learning Collatz - The Mother of all Rabbit Holes

What happens when you train a transformer on one of math's most infamous unsolved puzzles—and then study how it fails?