Models of Human Memory and Attention
Hayoung Song, PhD
Postdoctoral Research Fellow
Center for Theoretical & Computational Neuroscience (CTCN)
WashU
I will present two projects. The first introduces a normative model of human memory, showing how a recurrent neural network with episodic memory can learn the causal relationships between narrative events. The second takes a data-driven dynamical systems approach, showing how the geometry of neural dynamics along the cortical attractor landscape reflects changes in attention across time and context. Together, we will discuss how computational models can complement empirical findings to explain better the complex cognitive processes that shape our everyday experiences.