MATHEMATICS
Okay, plot twist: math and I met long before the country rankings. In elementary school, I read about the Seven Bridges of Königsberg, understood almost nothing, and still couldn’t sleep. That stubbornness suited me, because my ADHD meant I couldn’t focus for long on schoolwork, yet I could hyperfixate on a single topic for four hours straight. Math became the one place where that quirk was a superpower.
In middle school, during COVID, I taught myself math alone and at my own pace, and I started winning local competitions. Because I had overcome a challenge and still succeeded, I naively wondered whether everyone could do the same. Then I wondered about the classmates who were financially barred from private tutoring: could they ever reach the same merit? The question stayed with me, and I started asking it of everything around me. 👀
In high school, the flame began to flicker. I moved far from home, still refused extra tutoring, and my competition scores dropped as I felt myself falling behind. Math had been the puzzle I loved, and now it felt like a race I was losing to people who had started with a head start. So, I eventually decided to stop participate Olympiads.
I still love math, though it now speaks to me in a newer language. The puzzle I once solved for fun has become a tool for a harder question: how should a society redistribute opportunity? In economics, that question is a math problem, from optimal transport (moving resources at the lowest cost) to designing selection systems that don’t confuse a head start with talent. I no longer want my mathematics to win me medals. I want it to serve people, starting with the ones the old race left behind.

Mathematics and my earliest inquiries



GENERATIVE MODELS / PiMA
I became more interested in mathematics when it stopped feeling like a prepared problem and started becoming a way to describe something real. Optimal Transport was one example: a way to think about how resources move under cost and constraints. PiMA gave me another version of the same attraction—probability distributions unfolding over time to generate sound. Different domains, but the same thing kept pulling me in: hidden structure.
||S₄| = 4! = 242²⁴ candidate subsets → 310
Python + combinatorial arguments
OPTIMAL TRANSPORT / QPOT
ALGEBRAIC COMBINATORICS
MATHEMATICS BEYOND THE PROBLEM SET