Seminars

Can nonhuman animals perceive the relation-between-relations? This intriguing question has been studied over the last 40 years; nonetheless, the extent to which nonhuman species can do so remains controversial. Here, I review empirical evidence suggesting that pigeons, parrots, crows, and baboons join humans in reliably acquiring and transferring relational matching-to-sample (RMTS). Many theorists consider that RMTS captures the essence of analogy, because basic to analogy is appreciating the ‘relation between relations.’ Factors affecting RMTS performance include: prior training experience, the entropy of the sample stimulus, and whether the items that serve as sample stimuli can also serve as choice stimuli.

Whilst spatial navigation is a function ascribed to the hippocampus, flexibly adapting to a change in rule depends on the medial prefrontal cortex (mPFC). Single-units were recorded from the hippocampus and mPFC of rats shifting between a spatially- and cue-guided rule on a plus-maze. The mPFC population coded for the relative position between start and goal arm. During awake immobility periods, the mPFC replayed organized sequences of generalized positions which positively correlated with rule-switching performance. Conversely, hippocampal replay negatively correlated with performance and occurred independently of mPFC replay. Sequential replay in the hippocampus and mPFC may thus serve different functions.

One of the fundamental functions of the brain is to flexibly plan and control movement production at different timescales in order to efficiently shape structured behaviors. I will present research investigating how these complex computations are performed in the mammalian brain, with an emphasis on autonomous motor control. Specifically, I will focus on the mechanisms supporting efficient interfacing between 'higher-level' planning commands and 'lower-level' motor cortical dynamics that ultimately drive muscles. I will take advantage of the fact that the anatomy of the circuits underlying motor control is well known. It notably involves the primary motor cortex, a recurrent network that generates learned commands to drive muscles while interacting through loops with thalamic neurons that lack recurrent excitation. Using an analytically tractable model that incorporates these architectural constraints, I will explain how this motor circuit can implement a form of efficient modularity by combining (i) plastic thalamocortical loops that are movement-specific and (ii) shared hardwired circuits. I will show that this modular architecture can balance two different objectives: first, supporting the flexible recombination of an extensible library of re-usable motor primitives; and second, promoting the efficient use of neural resources by taking advantage of shared connections between modules. I will finally show that these insights are relevant for designing artificial neural networks able to flexibly and robustly compose hierarchical analog behaviors from a library of motor primitives. Presented in the van Vreeswijk Theoretical Neuroscience Seminar series (formerly WWTNS) on 2023-01-11. Recording duration: 00:49:52.

An influential theory in systems neuroscience suggests that brain function can be understood through low-dimensional dynamics [Vyas et al 2020]. However, a challenge in this framework is that a single computational task may involve a range of dynamic processes. To understand which processes are at play in the brain, it is important to use data on neural activity to constrain models. In this study, we present a method for extracting low-dimensional dynamics from data using low-rank recurrent neural networks (lrRNNs), a highly expressive and understandable type of model [Mastrogiuseppe & Ostojic 2018, Dubreuil, Valente et al. 2022]. We first test our approach using synthetic data created from full-rank RNNs that have been trained on various brain tasks. We find that lrRNNs fitted to neural activity allow us to identify the collective computational processes and make new predictions for inactivations in the original RNNs. We then apply our method to data recorded from the prefrontal cortex of primates during a context-dependent decision-making task. Our approach enables us to assign computational roles to the different latent variables and provides a mechanistic model of the recorded dynamics, which can be used to perform in silico experiments like inactivations and provide testable predictions.

Suppose you are on your favorite touchscreen device consciously and deliberately deciding emails to read or delete. In other words, you are consciously and intentionally looking, tapping, and swiping. Now suppose that you are doing this while neuroscientists are recording your brain activity. Eventually, the neuroscientists are familiar enough with your brain activity and behavior that they run an experiment with subliminal cues which reveals that your looking, tapping, and swiping seem to be determined by a random switch in your brain. You are not aware of it, or its impact on your decisions or movements. Would these predictions undermine your sense of free will? Some have argued that it should. Although this inference from unreflective and/or random intention mechanisms to free will skepticism, may seem intuitive at first, there are already objections to it. So, even if this thought experiment is plausible, it may not actually undermine our sense of free will.

Understanding Human ability to learn novel concepts from just a few sensory experiences is a fundamental problem in cognitive neuroscience. I will describe a recent work with Ben Sorcher and Surya Ganguli (PNAS, October 2022) in which we propose a simple, biologically plausible, and mathematically tractable neural mechanism for few-shot learning of naturalistic concepts. We posit that the concepts that can be learned from few examples are defined by tightly circumscribed manifolds in the neural firing-rate space of higher-order sensory areas. Discrimination between novel concepts is performed by downstream neurons implementing ‘prototype’ decision rule, in which a test example is classified according to the nearest prototype constructed from the few training examples. We show that prototype few-shot learning achieves high few-shot learning accuracy on natural visual concepts using both macaque inferotemporal cortex representations and deep neural network (DNN) models of these representations. We develop a mathematical theory that links few-shot learning to the geometric properties of the neural concept manifolds and demonstrate its agreement with our numerical simulations across different DNNs as well as different layers. Intriguingly, we observe striking mismatches between the geometry of manifolds in intermediate stages of the primate visual pathway and in trained DNNs. Finally, we show that linguistic descriptors of visual concepts can be used to discriminate images belonging to novel concepts, without any prior visual experience of these concepts (a task known as ‘zero-shot’ learning), indicated a remarkable alignment of manifold representations of concepts in visual and language modalities. I will discuss ongoing effort to extend this work to other high level cognitive tasks.

Developmental prosopagnosia is characterised by severe, lifelong difficulties when recognising facial identity. Most researchers require prosopagnosia cases exhibit ultra-conservative levels of impairment on the Cambridge Face Memory Test before they include them in their experiments. This results in the majority of people who believe that they have this condition being excluded from the scientific literature. In this talk I outline the many issues that will afflict prosopagnosia research if this continues, and show that these excluded cases do exhibit impairments on all commonly used diagnostic tests when a group-based method of assessment is utilised. I propose a paradigm shift away from cognitive task-based approaches to diagnosing prosopagnosia, and outline a new way that researchers can investigate this condition.

Genome-wide association studies and functional genomics studies have linked specific cell types, genes, and pathways to Alzheimer’s disease (AD) risk. In particular, AD risk alleles primarily affect the abundance or structure, and thus the activity, of genes expressed in macrophages, strongly implicating microglia (the brain-resident macrophages) in the etiology of AD. These genes converge on pathways (endocytosis/phagocytosis, cholesterol metabolism, and immune response) with critical roles in core macrophage functions such as efferocytosis. Here, we review these pathways, highlighting relevant genes identified in the latest AD genetics and genomics studies, and describe how they may contribute to AD pathogenesis. Investigating the functional impact of AD-associated variants and genes in microglia is essential for elucidating disease risk mechanisms and developing effective therapeutic approaches." https://doi.org/10.1016/j.neuron.2022.10.015

In the first part of this talk I will present work showing that recurrent neural networks can replicate broad behavioral patterns associated with dynamic visual object recognition in humans. An analysis of these networks shows that different types of recurrence use different strategies to solve the object recognition problem. The similarities between artificial neural networks and the brain presents another opportunity, beyond using them just as models of biological processing. In the second part of this talk, I will discuss—and solicit feedback on—a proposed research plan for testing a wide range of analysis tools frequently applied to neural data on artificial neural networks. I will present the motivation for this approach as well as the form the results could take and how this would benefit neuroscience.

The impact of apolipoprotein E ε4 (APOE4), the strongest genetic risk factor for Alzheimer’s disease (AD), on human brain cellular function remains unclear. Here, we investigated the effects of APOE4 on brain cell types derived from population and isogenic human induced pluripotent stem cells, post-mortem brain, and APOE targeted replacement mice. Population and isogenic models demonstrate that APOE4 local haplotype, rather than a single risk allele, contributes to risk. Global transcriptomic analyses reveal human-specific, APOE4-driven lipid metabolic dysregulation in astrocytes and microglia. APOE4 enhances de novo cholesterol synthesis despite elevated intracellular cholesterol due to lysosomal cholesterol sequestration in astrocytes. Further, matrisome dysregulation is associated with upregulated chemotaxis, glial activation, and lipid biosynthesis in astrocytes co-cultured with neurons, which recapitulates altered astrocyte matrisome signaling in human brain. Thus, APOE4 initiates glia-specific cell and non-cell autonomous dysregulation that may contribute to increased AD risk." https://doi.org/10.1016/j.cell.2022.05.017

Recent studies have begun to reveal critical roles for the brain’s professional phagocytes, microglia, and their receptors in the control of neurotoxic amyloid beta (Aβ) and myelin debris accumulation in neurodegenerative disease. However, the critical intracellular molecules that orchestrate neuroprotective functions of microglia remain poorly understood. In our studies, we find that targeted deletion of SYK in microglia leads to exacerbated Aβ deposition, aggravated neuropathology, and cognitive defects in the 5xFAD mouse model of Alzheimer’s disease (AD). Disruption of SYK signaling in this AD model was further shown to impede the development of disease-associated microglia (DAM), alter AKT/GSK3β-signaling, and restrict Aβ phagocytosis by microglia. Conversely, receptor-mediated activation of SYK limits Aβ load. We also found that SYK critically regulates microglial phagocytosis and DAM acquisition in demyelinating disease. Collectively, these results broaden our understanding of the key innate immune signaling molecules that instruct beneficial microglial functions in response to neurotoxic material." https://doi.org/10.1016/j.cell.2022.09.030