I am:

• a DPhil (PhD) student in the Reasoning with Machines Lab, supervised by Adam Mahdi at the Oxford Internet Institute;
• an inaugural Cosmos Fellow at the Laboratory for Human-Centered AI, supervised by Philipp Koralus and supported by the Cosmos Institute.

I research evaluations of AI systems. Evaluations are how we discover and characterise the space of intelligent behaviour in AI. I am concerned with building out a metrology for this evaluation practice (a scientific study of the measurement tools). So, you can think of my research as lens crafting, rather than astronomy—I build the “telescopes for intelligence” rather than the intelligent systems themselves.

My academic work addresses the space of questions like these:

• Do modern AI systems have cognitive capabilities?
• How can we build valid measures of intelligent behaviour in machines?
• How can we ascribe cognitive concepts to AI systems in an epistemologically and scientifically suitable way?
• How does (and should) our ascription of these cognitive concepts change the shape of our interactions with AI?

Previously, I studied Philosophy (BAH) and Computer Science (BS) at Stanford University. I wrote a thesis about trust—what it is, philosophically, and what is required for a philosophically defensible account of trust in explainable AI literature. More recently, I completed my MSc at the Oxford Internet Institute. My MSc thesis addresses the more technical question of construct validity in large language model evaluations (and should be publicly available soon!)

I was also a Founder in Residence at Entrepreneurs First, worked at Monte Carlo Data as a Founding Data Scientist, did NLP research with Stanford OVAL, and wrote three chapters of the Data Quality Fundamentals textbook with O’Reilly Media.

You can email me by decrypting this Caesar Cipher: ipXe%b\Xiej7f``%fo%XZ%lb