Michael Chen
Hi there! I'm an undergrad studying computer science at CMU. Among other things, I'm interested in AI, music, and baseball. Currently I'm working on a research project studying the effects of different pretraining choices on LLM performance.
Projects
Baseball batting simulator built in Unreal Engine 4. Simulates gravity, drag, and Magnus forces on baseball pitches. See code here
Exeter Course Map: a visual curriculum-planning tool which maps my school's 400+ academic courses in tree-styled graphs. See code here
202020: WPF Windows app to make the 20-20-20 eye health rule more convenient; >11k downloads and 4.7/5 rating. Microsoft store link
Dexcellence crypto-trading platform, built at the 72-hour ETHNewYork 2022 blockchain hackathon. Won 3 prizes.
Visual demonstration of A* pathfinding algorithm written in Processing Java, with 40 points and 45 edges (purple highlights indicate which paths were considered). See code here
Handwritten KMeans clustering algorithm, outperforming the Scikit-learn library implementation on a 111-MB dataset of Spotify songs.
Social music app built with Flutter and Firebase. Implements a "streak" system like Snapchat, where friends send each other songs from their recent Spotify history.
2Y2B: Made at CMU's TartanHacks hackathon. User inputs news sources/topics they follow, 2Y2B gives an-AI generated 5-minute news briefing. Built with Swift, Firebase, React, Flask, OpenAI API, and Google TTS.
Recent Stuff
What I've been reading/watching (will definitely forget to update this):
-
Shōgun
(2024)
10/10 -
Linear Algebra Done Right
should've read this before my linear algebra class last fall 😭 - A Brief History of Intelligence: Evolution, AI, and the Five Breakthroughs That Made Our Brains