Computational & Experimental Investigations of Learning & Memory
We routinely learn to go from a place 'A' to a neighboring 'B', often within a single trial, even in highly unfamiliar environments. We are interested in understanding how such maps of the world are learned and remembered. Learning is believed mediated by synaptic plasticity. Research in the past three decades has shown that synaptic plasticity depends critically on the relative spike timing of neurons. Do neuronal activity patterns, that are required for synaptic plasticity, occur during behavior? If so, do they induce synaptic plasticity? What are the effects of synaptic plasticity on the structure and activity of complex neuronal networks and behavior?
To address these questions, we record the activity of a large number of neurons from the hippocampus and related cortical regions during behavior. Hippocampal neurons fire selectively as a function of the animal's spatial location, and the hippocampal synapses show synaptic plasticity. Hence, we have set up computational models of learning spatial maps via synaptic plasticity. Our models suggest that oscillations are critical for generating the neuronal activity patterns required for synaptic plasticity. Further, these models predict that synaptic plasticity would make the hippocampal place fields more asymmetric and 'anticipatory' with experience, that could allow an animal to predict the upcoming spatial location based on past experience. Data from our lab, as well as from other labs, support these computational hypotheses. The results generalize easily to learning temporal sequences in other parts of the cortex as well, and pave the way towards a better understanding of mechanisms of learning and memory in neural networks.
Problems with the website? Please contact Omar Jamil Ahmed (omar_ahmed@brown.edu)