Shinya Ito
Scientist III
Shinya Ito is Scientist III in Anton Arkhipov’s team at the Allen Institute. He develops biologically realistic models of the mouse primary visual cortex, integrating electron-microscopy- and electrophysiology-derived connectivity with machine-learning optimization to reveal how circuit structure supports computation and to advance biologically grounded AI. Trained as a physicist (B.S., Hokkaido University), he earned a Ph.D. in Physics from Indiana University with John Beggs, mapping functional neuronal connectivity, and completed postdoctoral work at UC Santa Cruz with Alan Litke and David Feldheim, pioneering high-density silicon-probe recordings of the mouse superior colliculus to characterize its visual and auditory responses. Ito now combines large-scale simulations, TensorFlow-based training, and the Allen Institute’s in-house Brain Modeling ToolKit (BMTK), to which he is an active contributor, to build, train, and analyze realistic cortical networks.
research focus
As neuroscience is entering the era of big data, one of the goals in the field is to distill them into computational principles and mechanisms. Biophysically detailed modeling is potentially one of the ways to achieve the goal. Shinya’s focus is to take cutting-edge dataset in the field and extract underlying statistics and implement them in the model. The resulting model will bridge between the biophysical structures and the functional properties of the neuronal network and provide a laboratory for testing hypotheses of visual information processing.
