research

“All animals suffer unhappiness because learning is design, and this particular design for learning is optimal…” - from Principles of Neural Design (Sterling and Laughlin)

currently

I completed my Ph.D. in the Computational Cognitive Neuroscience Lab in May 2024, where I worked on computational theories of learning and decision making under cognitive resource constraints. I’m curious about how cognitive constraints can both benefit and limit our behavioral repertoire, and how the trade-off between reward and cognitive effort is compromised in psychiatric disease. I am also interested more broadly in the limitations of human reasoning and thought, and in the collaboration between artificial and natural intelligence. My thesis, Policy compression: Acting with limited cognitive resources, can be found here.

From September 2023, I am a TSVP Visiting Scholar with Gail Tripp and Kenji Doya at the beautiful Okinawa Institute of Science and Technology.

publications

Liu, S., Lai, L., Gershman, S.J., & Bari, B.A. (submitted). Time and memory costs jointly determine a speed-accuracy trade-off and set-size effects.
Lai, L., & Gershman, S.J. (2024). Human decision making balances reward maximization and policy compression. PLOS Computational Biology
Lai, L., Huang, A.Z.X., & Gershman, S.J. (submitted). Action chunking as policy compression.
Lai, L. & Gershman, S.J. (2021). Policy compression: an information bottleneck in action selection. Psychology of Learning and Motivation.
Gershman, S.J. & Lai, L. (2021). The reward-complexity trade-off in schizophrenia. Computational Psychiatry.
Bhui, R., Lai, L., Gershman, S.J. (2021). Resource-rational decision making. Current Opinion in Behavioral Sciences.
Mikhael, J.G., Lai, L., & Gershman, S.J. (2021). Rational inattention and tonic dopamine. PLOS Computational Biology.
Lai, L., Magnotti, J.F., & Yau, J.M. (2017). Multisensory context warps time perception. 2017 Conference on Cognitive Computational Neuroscience.

previously

My undergraduate research days were spent in the Yau Lab at Baylor College of Medicine. There I worked on developing a Bayesian inference model for understanding multisensory time distortions.

I often fled north to escape hot Houston summers… :-)

Thanks to an NSF-REU, I spent some time at MIT working in the lab of Mehrdad Jazayeri. My work there focused on understanding how variability in perception, inference, and memory influences sensorimotor timing.

I spent another summer at the beautiful Janelia Research Campus as a part of the JUS program. There, Josh Dudman taught me to record from neurons, and together we tried to understand how the striatum might represent action kinematics.

back