Understanding how the brain supports flexible behavior in dynamic environments is a key topic in cognitive neuroscience. In everyday life, we continuously adapt our thoughts and actions in response to changing goals, unexpected outcomes, and novel information. This ability—known as cognitive flexibility—is essential for survival and is often affected in psychiatric and neurological disorders. Despite its importance, the neural mechanisms underlying cognitive flexibility are not fully understood. To address this gap, this thesis focused on two key components of cognitive flexibility: decision making and learning. Using neural population analysis and network modeling on decision-making and reversal learning tasks, this work examined how adaptive behavior was supported by distributed computations across PFC subregions and striatum, with a primary emphasis on the PFC. Combined with a synthesis of recent circuit models, this thesis aimed to provide a mechanistic understanding of how cognitive flexibility is implemented in the brain.
Xiaochen Zhao was born on December 25th,1992, in Rushan, China, where she finished her high school in 2011. She then moved to Chengdu, China, where she studied biomedical engineering at the University of Electronic Science and Technology of China. In 2015, she completed her Bachelor’s thesis under the supervision of Dr. Malte Rasch. Her Bachelor’s thesis was about exploring the neural mechanisms of visual perceptual learning by training a multi-layer neural network to perform a contour detection task with reinforcement learning. After that, she pursued her Master’s degree at the National Key Laboratory of Cognitive Neuroscience and Learning in Beijing Normal University (Beijing, China). In 2018, she finished her master’s thesis under the supervision of Dr. Malte Rasch and Prof. Si Wu. The thesis investigated the spatial-temporal dynamics of information propagation in the primary visual cortex (V1) by combining electrophysiological data analysis and spiking neural network models. In the same year, she became a PhD student at Donders Institute for Brain, Cognition and Behaviour in Radboud University (Nijmegen, the Netherlands), under the supervision of Prof. Paul Tiesinga and Prof. Thilo Womelsdorf, to work on the neural basis of flexible behavior using a combination of tools, including neural population analysis, reinforcement learning, and neural network models.