I'm a computational cognitive scientist finishing my PhD between Harvard and MIT as a Harvard Prize Fellow. I largely care about understanding human intelligence, both for its own sake and as a means for building smart machines that are more like people. I've worked as a research scientist at the MIT-IBM Watson lab developing tractable models of commonsense reasoning of the physical world and Google Deepmind building evaluations of compositional generalization for Gemini and co-leading an internal project on getting machines to write their own software. Most recently, I've founded the world's first cognitive science frontier lab with the explicit goal of reverse-engineering human intelligence as a path toward AGI.
