万烨鹏
I write code and play ranked.
I’m Wilson. I studied CS at Waterloo and am now at UIUC. My main interests are quant research and development, and I like exploring different areas of AI.
32 random paths. Pick only the best one, and randomness starts looking like a strategy.
32 paths. Each step has an equal chance of going up or down.
A small experiment in selection bias. No market data or predictions.
AlphaForge
A quant research tool for daily data, covering universe selection, signals, trading costs, and reporting. In a ten-day synthetic example, the strategy returned 1.38% against the benchmark’s 3.34%. The curve goes up, but it still trails the benchmark.
AI Commerce Agent
“Like this, but better for running.”
Search combines text, images, and keywords, then reranks the results around what the user wants. In this saved example, a request for running shoes first returns a formal shoe. Rules and model reranking then move a sports shoe to the top.
What I’m working on, what I’m learning, and a little of everything else.
I started at Waterloo in 2020, so my first experience of university was online. I eventually made it to campus in Canada, then moved on to UIUC in the US.
My internships have involved wireless-device testing, skate-blade profiles, battery charging, and quant development. This site is where I share projects, write things down, and talk a little about games.
Read the storyI’ve spent plenty of time in CrossFire, CS:GO, and League.
About a project, an idea, or a good game.
