Maria Ryskina she/they

I am a CIFAR AI Safety Postdoctoral Fellow at the Vector Institute for AI. I work with Gillian Hadfield on interdisciplinary approaches to AI alignment, building on studies of how social norms and rules are shaped by human societies. My research interests are diverse, spanning artificial intelligence, natural language processing, linguistics, and cognitive and behavioural sciences.

Previously, I was a postdoc at Ev Fedorenko’s lab at MIT, studying how concepts are represented in the brain and in large language models. Earlier, I completed my PhD in Language and Information Technologies at Carnegie Mellon University, building computational models of non-standard language uses (with Taylor Berg-Kirkpatrick, Matt Gormley, and Ed Hovy). I recieved my BSc and MSc degrees from the Moscow Institute of Physics and Technology and Skoltech. I had also visited MIT InfoLab (2015) and interned at DiDi Labs (2020).

Outside of research, I am a core organizer at Queer in AI and serve on the board of the Toronto ACM-W chapter (more info coming soon!)