The brain's remarkable characteristic lies in its ability to process information and control behavior through networks of diverse neurons. We are dedicated to exploring the quantitative principles of biological neural systems at the network level, focusing on the following areas:

Neural Coding

The nervous system represents and processes information through the collective activity of neurons. Constrained by both internal and external environments, biological neural coding exhibits unique yet universal characteristics. We study the formation, transformation, and development of olfactory neural coding in fruit flies, exploring the fundamental principles of biological neural coding from perspectives such as statistics, dynamics, and neural computation. These principles are then applied to the development of brain-inspired algorithms and bionic systems.

Circuit Assembly

The structure of neural networks largely determines the properties and functions of the nervous system. Developmental processes, physical constraints, and neural activity all leave imprints on network structure. By analyzing the connectomes of organisms likeC. elegansand fruit flies, and tracking their development and plasticity, we identify key factors influencing network structure across multiple spatiotemporal scales and construct dynamic models of neural circuit assembly.

Motor Control

Motor control is one of the most crucial functions of the nervous system, with information processing ultimately needing to manifest in behavior. The uniqueness of movement lies in its being the result of interactions among the nervous system, body and environment. We combine mechanics, control theory, neuroscience, and virtual reality to study the movement of fruit fly larvae, exploring the fundamental principles of biological motor control to inspire new approaches in the design and development of bionic robots.