Ali Behrouz
Staff Research Scientist · Google
Final-year Ph.D. Candidate, Computer Science · Cornell University
I am a Staff Research Scientist at Google and a final-year Ph.D. candidate in Computer Science at Cornell University. My research interests broadly span machine learning methods and understanding intelligence. These days, I mainly work on designing deep learning architectures — models with long-term memory that learn to memorize at test time, retain and consolidate knowledge, and learn continually. This line of work includes Titans, ATLAS, Miras, and Nested Learning (Hope).
Before joining Cornell, I completed my M.Sc. in Computer Science at the University of British Columbia, where I was fortunate to be advised by Prof. Margo Seltzer. Before that, I received my B.Sc. in Computer Engineering with a minor in Mathematics from Sharif University of Technology.
I am always open to collaboration!
I would be happy to collaborate with Ph.D., Master’s, and Bachelor’s students who would like to work on machine learning methods and deep learning architectures. I also have some projects in these areas that we can discuss. Feel free to reach out by email, and please include “[Collaboration]” in the subject line.
Selected Publications
-
Nested Learning: The Illusion of Deep Learning Architectures
-
Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories
-
ATLAS: Learning to Optimally Memorize the Context at Test Time
See the full publication list or my Google Scholar profile.