Learning to think by reinforcing useful thoughts
WTI Inspiring Speaker: Timothy Lillicrap
Thursday, October 9, 2025
10:00 - 11:15 am
100 College Street
Floor 11, Workshop 1116
The ability to spend time and compute to produce better answers is a hallmark of human cognition. But how can we teach machines to think? Timothy Lillicrap's talk will focus on how reinforcement learning algorithms can produce models that reason deeply about everything from games to math, and everything else we think about. Shreya Saxena will be the host for Lillicrap's visit.
Timothy Lillicrap
Research Director, Google DeepMind
Lillicrap's research centers on machine learning and statistics for optimal control and decision-making, developing new algorithms and approaches to using deep networks in reinforcement learning, and novel network and memory architectures. More recently, his work focuses on developing reinforcement learning methods for teaching reasoning in large language models.
WTI Inspiring Speaker Series
The Wu Tsai Institute's Inspiring Speaker Series features an interdisciplinary lineup of speakers whose work reshapes how we understand the mind and brain. Talks occur monthly on Thursdays at 10:00 am, from September 2025 to May 2026, at 100 College Street. Coffee and refreshments will be available before each talk, and there will be an opportunity for questions afterward. All members of the Yale community are welcome.
If you would like to meet with a visiting speaker, please contact wti@yale.edu.