Prof. FUNG Chi Chung Alan (馮志聰)

Dr. Alan Chi Chung Fung obtained his Ph.D. degree in Physics from the Hong Kong University of Science and Technology (HKUST) in 2013. In 2015 – 2019, he served as a research scientist at the RIKEN Centre of Brain Science (CBS). After that, he worked as a staff scientist at the Okinawa Institute of Science and Technology Graduate University (OIST) until January 2022. In late February 2022, he joined the City University of Hong Kong.

Research Contribution

Alan started his research on continuous-attractor neural network (CANN) models in 2006. He has jointly discovered the intrinsic dynamics of CANN under the influence of dynamical synapses (Fung, Wong & Wu, 2010; Fung et al., 2012). After that, he has jointly proposed the anticipatory behavior observed in CANN. The anticipatory behavior can explain how the head-direction (HD) cells in the rat anterior thalamic nucleus (ATN) encode future head directions (Fung, Wong & Wu, 2012). Recently, Alan found that the CANN with short-term depression can address the discrete-attractor-like behavior observed in the hippocampus during periods of non-rapid-eye-movement sleep (Fung & Fukai, 2019).

Besides CANN, Alan has also contributed other modeling works to address phenomena observed in experiments. In 2016, Alan jointly proposed a mathematical model to explain the integration of saccadic motion and visual signal (Wang et al., 2016). In 2019, Alan jointly contributed an algorithm based on adult neurogenesis observed in the dentate gyrus to perform pattern separation. The algorithm can be a potential machine learning algorithm for future usage. In 2021, Alan supplied a series of predictions on long-term synaptic plasticity due to neurotransmitter release probability distribution modulated by the suppression of NMDA astrocytic receptors (Chipman et al., 2021). The further implication is still under investigation.

Alan's current research interests include (1) mathematical models for neural phenomena, (2) neuroscience-inspired algorithms, and (3) analyzing neural data. Mathematical models enable scientists to explore broader conditions of neural phenomena and potential implications. By developing algorithms based on neural phenomena, one may retrieve the corresponding functional meaning in information processing. On the other hand, it will be interesting to see how modern data science skills may unveil discoveries from existing neural data.