Tensor decomposition

Over the past decade, neural representations have been studied from the lens of low-dimensional subspaces defined by the co-activation of neurons. However, this view has overlooked other forms of covarying structure in neural activity, including i) condition-specific high-dimensional neural sequences, and ii) representations that change over time due to learning or drift. In this talk, I will present a new framework that extends the classic view towards additional types of covariability that are not constrained to a fixed, low-dimensional subspace. In addition, I will present sliceTCA, a new t

Over the past decade, neural representations have been studied from the lens of low-dimensional subspaces defined by the co-activation of neurons. However, this view has overlooked other forms of covarying structure in neural activity, including i) condition-specific high-dimensional neural sequences, and ii) representations that change over time due to learning or drift. In this talk, I will present a new framework that extends the classic view towards additional types of covariability that are not constrained to a fixed, low-dimensional subspace. In addition, I will present sliceTCA, a new t

Mice are territorial animals whose spatial navigation ability relies on integrating multiple stimuli from various sensory modalities. Among all these types of information, the pattern of major urinary proteins contained in the urine landmarks provide the identity of the conspecific owner of each territory, constituting the “who” component of episodic memory. We identified a neuroanatomical pathway responsible for integrating individual recognition information into the complex socio-spatial memories stored within the hippocampus. Within this pathway, the posteromedial cortical amygdaloid nucleu

…data into a three-dimensional array or tensor, of size neurons by time points by trials. Classic dimensionality reduction methods often assume that neural representations are constrained to a fixed low-dimensional latent subspace. Consequently, this view does not capture how the latent subspace could evolve over learning, nor how high-dimensional neural activity could emerge over learning. Furthermore, the link between these empirically-observed changes in neural

Exploiting invariances in the inputs is crucial for constructing efficient representations and accurate predictions in neural circuits. In neuroscience, translation invariance is at the heart of models of the visual system, while convolutional neural networks designed to exploit translation invariance triggered the first wave of deep learning successes. While the hallmarks of convolutions, namely localised receptive fields that tile the input space, can be implemented with fully-connected neural networks, learning convolutions directly from inputs in a fully-connected network has so far proven

During prolonged periods of stress, animals switch from active to passive coping strategies to manage effort expenditure. These normally adaptive behavioral state transitions become maladaptive in disorders such as depression. Such behavioral state transitions occur as animals continually evaluate the fruitfulness or futility of their actions in response to stressors. This process is mediated by interactions across numerous brain regions over long timespans, integrating information about the stressor with the outcome of avoidance actions performed, and tracking the accrual of stress, ultimatel