Speaker: Shanshan Qin Harvard University Cambridge, MA Title: Unveiling the dynamics and structure of drifting neural representations Abstract: While it was naturally assumed that stable behavior and long-term memories are associated with stable neuronal representations, recent long-term recordings have surprisingly shown that neuronal representations in brain regions associated with stable behaviors and memories continuously change in time scales of days and weeks, even after a task is fully learned. Key questions about the causes, dynamics, and functions of this “representational drift” remain unanswered. In this talk, I will show you our contributions to address these questions in a biologically plausible mechanistic model. We explore the hypothesis that neural representations optimize a representational objective with a degenerate solution space, and noisy synaptic updates drive the network to explore this (near-)optimal space causing representational drift. We illustrate this idea in simple, biologically plausible Hebbian/anti-Hebbian network models of representation learning, which optimize similarity matching objectives, and, when neural outputs are constrained to be nonnegative, learn localized receptive fields (RFs) that tile the stimulus manifold. We find that the drifting RFs of individual neurons can be characterized by a coordinated random walk, with the effective diffusion constants depending on various parameters such as learning rate, noise amplitude, and input statistics. Despite such drift, the representational similarity of population codes is stable over time. Our model recapitulates recent experimental observations in hippocampus and posterior parietal cortex, and makes testable predictions that can be probed in future experiments.