Hi! I'm Darius. I'm interested in how advanced AI systems can interact with institutions, governance, and collective decision-making to advance social good. Recently I've been studying AI-delegated deliberation with the Cooperative AI Foundation and Pivotal Research.

I've previously researched chain-of-thought interpretability with the Cambridge AI Alignment Initiative and worked as a research engineer at Cartesia, where I trained and optimized large-scale text-to-speech models. I've also worked on large language model systems and research engineering through internships at NVIDIA and Microsoft.

Outside of research, I like thrifting clothes, reading fiction, and being a flâneur.