About

I'm a research engineer with a background in pure mathematics, now working on problems in AI and machine learning. I care about understanding what neural networks actually learn — mechanistically, not just empirically.

Background

My path into ML came through mathematics. I spent years thinking carefully about structure and abstraction, which turns out to be surprisingly useful for understanding why certain architectures work and others don't.

Research Interests

• Mechanistic interpretability
• Representation learning and geometry
• Scientific machine learning
• WebAssembly and in-browser ML inference

Currently

Open to new opportunities. If you're working on something interesting at the intersection of theory and practice in ML, I'd love to talk.

Contact

Find me on GitHub, LinkedIn, or reach out at hello@noahf.ai.