∇
γ
π
∂
Stephen Yang
Stanford, CA
I build RL environments for software engineering agents at Mechanize. On leave from Stanford (Math & CS, '29). Before that, I wrote papers using LLMs to measure innovation quality and patenting disparities.
Currently
Software Engineer at Mechanize
Building RL environments for SWE agents
Say hello
yang29@stanford.edu
Papers
Prior research at the intersection of AI and economics.
Working Paper·2025
Understanding Innovation Quality and Success with Large-Language Models
Stephen Yang
Working Paper·2025
Understanding Patenting Disparities via Causal Human+Machine Learning
Lin William Cong, Stephen Yang
Working Paper·2024
Endogenous Fracturing under Partisan Voting
Baozhong Yang, Stephen Yang