January 31, 2017

Raimondo JV, Richards BA, Woodin MA, “Neuronal chloride and excitability — the big impact of small changes”, Current Opinion in Neurobiology, 43, 35-42

Synaptic inhibition is a critical regulator of neuronal excitability, and in the mature brain the majority of synaptic inhibition is mediated by Cl−-permeable GABAA receptors. Unlike other physiologically relevant ions, Cl− is dynamically regulated, and alterations in the Cl−gradient can have significant impact on neuronal excitability. Due to changes in the neuronal Cl− concentration, GABAergic transmission can bidirectionally regulate the induction of excitatory synaptic plasticity and gate the closing of the critical period for monocular deprivation in visual cortex. GABAergic circuitry can also provide a powerful restraining mechanism for the spread of excitation, however Cl− extrusion mechanisms can become overwhelmed […]
January 31, 2017

Santoro A, Frankland PW, Richards BA, “Memory Transformation Enhances Reinforcement Learning in Dynamic Environments”, Journal of Neuroscience, 36 (48), 12228-12242

Over the course of systems consolidation, there is a switch from a reliance on detailed episodic memories to generalized schematic memories. This switch is sometimes referred to as “memory transformation.” Here we demonstrate a previously unappreciated benefit of memory transformation, namely, its ability to enhance reinforcement learning in a dynamic environment. We developed a neural network that is trained to find rewards in a foraging task where reward locations are continuously changing. The network can use memories for specific locations (episodic memories) and statistical patterns of locations (schematic memories) to guide its search. We find that switching from an episodic […]
January 31, 2017

Guerguiev J, Lillicrap TP, Richards BA, “Deep learning with segregated dendrites”, arXiv:1610.00161

Deep learning has led to significant advances in artificial intelligence, in part, by adopting strategies motivated by neurophysiology. However, it is unclear whether deep learning could occur in the real brain. Here, we show that deep learning can be achieved by moving away from point neuron models and towards multi-compartment neurons. Like neocortical pyramidal neurons, neurons in our model receive feedforward sensory information and higher-order feedback in electrotonically segregated compartments. Thanks to this segregation, the network can calculate local synaptic weight updates that allow it to categorize images from the MNIST data-set with good accuracy. We show that our learning […]