How can a system learn from experience, retain what matters, and apply it in an unfamiliar situation? We’re interested in adaptation that continues beyond an initial training process.
The questions ahead
How far canintelligence go?
Learning, reasoning, and the foundations of intelligence. A long view, beginning with better questions.
Explore researchWe’re interested in how intelligent systems learn, reason, remember, and act, and how those abilities can become useful beyond a single task. These questions define our research direction.
01 / Research interests
The questions
that pull us forward.
Planning, reasoning, and acting are closely connected. Understanding that connection means asking how systems can break down complex tasks, check their own work, and recognize uncertainty.
What allows a capability to transfer across tasks and environments? Foundational models, architectures, and ways of evaluating progress all have a part in exploring that question.
02 / The long view
A wider
horizon.
Our long-term ambition includes foundational work toward artificial general intelligence and, beyond it, superintelligence. An ambitious horizon calls for sustained inquiry, careful evaluation, and the freedom to follow the evidence.
A research direction. A great deal still to discover.
An open invitation
A question worth exploring together?
We’re open to thoughtful research conversations and collaboration.