Discovering principles of flexible information processing in the cortex is critical to understanding the mechanisms of brain-state dependent sensory processing and cognition. We investigate the principles of flexible computations in multiple neural substrates: synaptic integration in single neurons, dynamics in excitatory-inhibitory networks, information routing between networks. In combination with theoretical frameworks such as dynamical systems and information theory, we use tools such as probabilistic graphical models, unsupervised learning, multicompartmental biophysical models and phenomenological models to conduct our research. Work in our group ranges from purely computational to active collaborations with experimental groups. ​