hey, i’m lydia!


currently i’m researching multi-agent reinforcement learning at the foerster lab. next year i get to study maths, all maths, for the first time in a long time! my lab project’s on meta-learning algorithms for cooperation in general-sum games.

i’m currently learning more about cognitive enhancement and r&d automation. i did some theoretical ai safety projects in the past (unilateralist’s curse, inter-ai mediation, swiss cheese approach for decorrelating failure modes).

communities i’m part of: fabric’s espr/warp, ev, ef’s polaris, ox’s merton. in past lives i studied ppe, worked at ftx, was a climate activist, and made education a force for peace at uwc hong kong (not in that order).