Arshad Kazi
exploring “brain”
If you're reading this, you probably ended up here by accident.
I am a trained engineer working around machine learning. I build things mostly to understand how they work, not just to make them work. Most of my work involves implementing models from scratch and working through training dynamics rather than relying on prebuilt abstractions (see my work on GitHub).
I am not a researcher and not a conventional developer, even though I spend a lot of time building. I tend to follow problems that require some level of reasoning, and I usually end up exploring them by building, breaking, and figuring out what is actually going on.
My primary interest is computer vision, especially problems involving structure, geometry, and representation. Things that can be tested, visualized, and taken apart: 3D reconstruction, pose estimation, representation learning. Some of this lives in my notebooks.
I am intense when I am interested, atheist, and a chronic overthinker. Anxious, often unsatisfied, hard to relax. I jump between ideas, go deep into one, drop it halfway, and return later with a better understanding.
I have millions of interests, thousands of questions, and billions of problems I want to resolve. I think about color theory, neuroscience, food, film, music, language, and whatever else pulls me in on a given week. I write some of it down in my thoughts. Most of it stays in my head. I will stay strong though.
I am still trying to understand where I stand and what I am trying to do in the long run. This is where I dump things I am working through. Not optimized, not curated, and not meant for social media.
Still trying to get found. Maybe that and finding myself are the same search. IDK.
Yours truly,
Arshad
* This site is managed by AI, but the content written here is purely human.