In his research, Arseny combines the insights from Cognitive Science and Machine Learning, aiming to benefit both fields. Arseny will be working with Melanie Mitchell on the problem of measuring abstraction and analogy-making capabilities of AI systems.

More generally, he is interested in modeling how language and abstraction allow humans to communicate and share knowledge with each other. His dream is to live to see the day when NLP models can have meaningful conversations, "meaningful" defined simply as being able to learn and change their long-term beliefs not only during initial training but also through natural dialogue. To make a step in that direction, Arseny uses a range of different methods, from behavioral studies and emergent communication simulations that help us better understand the role of language in knowledge communication to developing new NLP model architectures that can support such "on the fly" learning.

A special place in Arseny's heart is reserved for online education. His passion for Machine Learning began with an online course by Andrew Ng, and he has since learned numerous skills in that way. Willing to give back to the free online education community, Arseny made one course himself (the course on Neural Networks on stepic.org [in russian]). Lastly, Arseny continues to work on his mathematical skills on a daily basis and encourages anyone to contact him if they'd like to organize a self-study or problem-solving group on one topic or another.

Arseny received a B.Sc. in Psychology and an M.Sc. in Neuroscience from Saint Petersburg University, while also completing a two-year program in Machine Learning and Software development at the Computer Science Center. Lastly, he received an M.Sc. in statistics and a Ph.D. in Cognitive Science from UC Irvine, under the guidance of Mark Steyvers.

Joe Bak-Coleman

Anthony Eagan

Jan Eissfeldt

Tara Fjellman

William Hynes

Gulce Kardes

Penny Mealy

Sodiq Mojeed

Arseny Moskvichev

Abhishek Yadav

Emma Zajdela