The central theme of my lab is focused on computational neuroscience and brain-inspired AI. We use computational models to decipher the biological strategies that the brain employs to compute and learn efficiently, robustly, and continually. Additionally, we draw inspiration from neuronal dynamics and structural motifs of the brain to develop new AI models aimed at addressing existing issues in deep neural networks (DNNs).

In recent years, our focus has particularly centered on the cerebellum, which has traditionally been explored mainly for its role in motor control within a supervised learning framework. However, recent data supports its involvement in anomaly detection, decision-making, cognition, and social abilities. We need a new theoretical platform to account for the multifunctionality of the cerebellum, given its relatively uniform structure.

While most inspirations for biological AI have come from the cerebral cortex, we are also interested in how the computational principles of the cerebellum can enhance existing AI models.