Shallow neural networks trained to detect collisions recover features of visual loom-selective neurons

Animals have evolved sophisticated visual circuits to solve a vital inference problem: detecting whether or not a visual signal corresponds to an object on a collision course. Such events are detected by specific circuits sensitive to visual looming, or objects increasing in size. Various computational models have been developed for these circuits, but how the collision-detection inference problem itself shapes the computational structures of these circuits remains unknown. Here, inspired by the distinctive structures of LPLC2 neurons in the visual system of Drosophila, we build anatomically-constrained shallow neural network models and train them to identify visual signals that correspond to impending collisions. Surprisingly, the optimization arrives at two distinct, opposing solutions, only one of which matches the actual dendritic weighting of LPLC2 neurons. Both solutions can solve the inference problem with high accuracy when the population size is large enough. The LPLC2-like solutions reproduces experimentally observed LPLC2 neuron responses for many stimuli, and reproduces canonical tuning of loom sensitive neurons, even though the models are never trained on neural data. Thus, LPLC2 neuron properties and tuning are predicted by optimizing an anatomically-constrained neural network to detect impending collisions. More generally, these results illustrate how optimizing inference tasks that are important for an animal's perceptual goals can reveal and explain computational properties of specific sensory neurons.

Corticothalamic Pathways From Layer 5: Emerging Roles in Computation and Pathology

Large portions of the thalamus receive strong driving input from cortical layer 5 (L5) neurons but the role of this important pathway in cortical and thalamic computations is not well understood. L5-recipient “higher-order” thalamic regions participate in cortico-thalamo-cortical (CTC) circuits that are increasingly recognized to be (1) anatomically and functionally distinct from better-studied “first-order” CTC networks, and (2) integral to cortical activity related to learning and perception. Additionally, studies are beginning to elucidate the clinical relevance of these networks, as dysfunction across these pathways have been implicated in several pathological states. In this review, we highlight recent advances in understanding L5 CTC networks across sensory modalities and brain regions, particularly studies leveraging cell-type-specific tools that allow precise experimental access to L5 CTC circuits. We aim to provide a focused and accessible summary of the anatomical, physiological, and computational properties of L5-originating CTC networks, and outline their underappreciated contribution in pathology. We particularly seek to connect single-neuron and synaptic properties to network (dys)function and emerging theories of cortical computation, and highlight information processing in L5 CTC networks as a promising focus for computational studies.

An Algorithm for Linearizing the Collatz Convergence

The Collatz dynamic is known to generate a complex quiver of sequences over natural numbers for which the inflation propensity remains so unpredictable it could be used to generate reliable proof-of-work algorithms for the cryptocurrency industry; it has so far resisted every attempt at linearizing its behavior. Here, we establish an ad hoc equivalent of modular arithmetics for Collatz sequences based on five arithmetic rules that we prove apply to the entire Collatz dynamical system and for which the iterations exactly define the full basin of attractions leading to any odd number. We further simulate these rules to gain insight into their quiver geometry and computational properties and observe that they linearize the proof of convergence of the full rows of the binary tree over odd numbers in their natural order, a result which, along with the full description of the basin of any odd number, has never been achieved before. We then provide two theoretical programs to explain why the five rules linearize Collatz convergence, one specifically dependent upon the Axiom of Choice and one on Peano arithmetic.

Reasoning on with Defeasibility in ASP

Reasoning on defeasible knowledge is a topic of interest in the area of description logics, as it is related to the need of representing exceptional instances in knowledge bases. In this direction, in our previous works we presented a framework for representing (contextualized) OWL RL knowledge bases with a notion of justified exceptions on defeasible axioms: reasoning in such framework is realized by a translation into ASP programs. The resulting reasoning process for OWL RL, however, introduces a complex encoding in order to capture reasoning on the negative information needed for reasoning on exceptions. In this paper, we apply the justified exception approach to knowledge bases in , that is, the language underlying OWL QL. We provide a definition for knowledge bases with defeasible axioms and study their semantic and computational properties. In particular, we study the effects of exceptions over unnamed individuals. The limited form of axioms allows us to formulate a simpler ASP encoding, where reasoning on negative information is managed by direct rules. The resulting materialization method gives rise to a complete reasoning procedure for instance checking in with defeasible axioms.1

Choice Logics and Their Computational Properties

Qualitative Choice Logic (QCL) and Conjunctive Choice Logic (CCL) are formalisms for preference handling, with especially QCL being well established in the field of AI. So far, analyses of these logics need to be done on a case-by-case basis, albeit they share several common features. This calls for a more general choice logic framework, with QCL and CCL as well as some of their derivatives being particular instantiations. We provide such a framework, which allows us, on the one hand, to easily define new choice logics and, on the other hand, to examine properties of different choice logics in a uniform setting. In particular, we investigate strong equivalence, a core concept in non-classical logics for understanding formula simplification, and computational complexity. Our analysis also yields new results for QCL and CCL. For example, we show that the main reasoning task regarding preferred models is ϴ₂P-complete for QCL and CCL, while being Δ₂P-complete for a newly introduced choice logic.

Interrogating theoretical models of neural computation with emergent property inference

A cornerstone of theoretical neuroscience is the circuit model: a system of equations that captures a hypothesized neural mechanism. Such models are valuable when they give rise to an experimentally observed phenomenon -- whether behavioral or a pattern of neural activity -- and thus can offer insights into neural computation. The operation of these circuits, like all models, critically depends on the choice of model parameters. A key step is then to identify the model parameters consistent with observed phenomena: to solve the inverse problem. In this work, we present a novel technique, emergent property inference (EPI), that brings the modern probabilistic modeling toolkit to theoretical neuroscience. When theorizing circuit models, theoreticians predominantly focus on reproducing computational properties rather than a particular dataset. Our method uses deep neural networks to learn parameter distributions with these computational properties. This methodology is introduced through a motivational example of parameter inference in the stomatogastric ganglion. EPI is then shown to allow precise control over the behavior of inferred parameters and to scale in parameter dimension better than alternative techniques. In the remainder of this work, we present novel theoretical findings in models of primary visual cortex and superior colliculus, which were gained through the examination of complex parametric structure captured by EPI. Beyond its scientific contribution, this work illustrates the variety of analyses possible once deep learning is harnessed towards solving theoretical inverse problems.

Generative replay for compositional visual understanding in the prefrontal-hippocampal circuit

Understanding the visual world is a constructive process. Whilst a frontal-hippocampal circuit is known to be essential for this task, little is known about the associated neuronal computations. Visual understanding appears superficially distinct from other known functions of this circuit, such as spatial reasoning and model-based planning, but recent models suggest deeper computational similarities. Here, using fMRI, we show that representations of a simple visual scene in these brain regions are relational and compositional - key computational properties theorised to support rapid construction of hippocampal maps. Using MEG, we show that rapid sequences of representations, akin to replay in spatial navigation and planning problems, are also engaged in visual construction. Whilst these sequences have previously been proposed as mechanisms to plan possible futures or learn from the past, here they are used to understand the present. Replay sequences form constructive hypotheses about possible scene configurations. These hypotheses play out in an optimal order for relational inference, progressing from predictable to uncertain scene elements, gradually constraining possible configurations, and converging on the correct scene configuration. Together, these results suggest a computational bridge between apparently distinct functions of hippocampal-prefrontal circuitry, and a role for generative replay in constructive inference and hypothesis testing.

The cell biology of synapse formation

In a neural circuit, synapses transfer information rapidly between neurons and transform this information during transfer. The diverse computational properties of synapses are shaped by the interactions between pre- and postsynaptic neurons. How synapses are assembled to form a neural circuit, and how the specificity of synaptic connections is achieved, is largely unknown. Here, I posit that synaptic adhesion molecules (SAMs) organize synapse formation. Diverse SAMs collaborate to achieve the astounding specificity and plasticity of synapses, with each SAM contributing different facets. In orchestrating synapse assembly, SAMs likely act as signal transduction devices. Although many candidate SAMs are known, only a few SAMs appear to have a major impact on synapse formation. Thus, a limited set of collaborating SAMs likely suffices to account for synapse formation. Strikingly, several SAMs are genetically linked to neuropsychiatric disorders, suggesting that impairments in synapse assembly are instrumental in the pathogenesis of neuropsychiatric disorders.

A taxonomy of the brain’s white matter: Twenty-one major tracts for the twenty-first century

The functional and computational properties of brain areas are determined, in large part, by their connectivity profiles. Advances in neuroimaging and network neuroscience allow us to characterize the human brain noninvasively and in vivo, but a comprehensive understanding of the human brain demands an account of the anatomy of brain connections. Long-range anatomical connections are instantiated by white matter and organized into tracts. Here, we aim to characterize the connections, morphology, traversal, and functions of the major white matter tracts in the brain. It is clear that there are significant discrepancies across different accounts of white matter tract anatomy, hindering our attempts to accurately map the connectivity of the human brain. We thoroughly synthesize accounts from multiple methods, but especially nonhuman primate tract-tracing and human diffusion tractography. Ultimately, we suggest that our synthesis provides an essential reference for neuroscientists and clinicians interested in brain connectivity and anatomy, allowing for the study of the association of white matter’s macro and microstructural properties with behavior, development, and disordered processes.