marcioodiaz.com

Marcio Diaz’s site presents his work as a mechanistic interpretability researcher and software engineer, highlighting his focus on reverse-engineering neural networks to understand their internal algorithms. It positions him as a researcher and engineer exploring AI safety and model transparency, with accessible blog content and research notes for readers interested in AI alignment and mechanistic insights.