efficiency

Cartilage targeting exosomes for OA gene therapy and pain treatment

Project Summary Gene therapy has the potential to facilitate targeted expression of therapeutic proteins to promote cartilage regeneration in osteoarthritis (OA). The dense, avascular, aggrecan-glycosaminoglycan rich negatively charged cartilage, however, hinders their transport to reach chondrocytes in effective doses. While viral vector mediated gene delivery has shown promise, concerns over immunogenicity and tumorigenic side-effects persist. To address this, we have developed surface-modified cartilage-targeting MSC exosomes as non-viral carriers for gene therapy. MSC derived exosomes have intrinsic therapeutic potential as they can induce cartilage repair and are non-immunogenic, making them desirable for gene delivery. We have engineered charge-reversed cationic exosomes by anchoring cartilage targeting optimally charged arginine-rich cationic peptide (CPC) motifs into the anionic exosome bilayer (Exo-CPC) by using buffer pH as a charge-reversal switch. Exo-CPC use charge interactions to penetrate through the full thickness of arthritic cartilage (close to tidemark) and deliver the packaged genetic material cargo to chondrocytes residing in the deep tissue layers while native anionic exosomes cannot. They can also bind within the synovial joint, making them effective for OA pain relief gene therapy. Here we will engineer charge-reversed Exo-CPC for delivery of IL-1RA (receptor antagonist of interleukin-1) mRNA and NaV1.8 (voltage gated sodium channel 1.8) inhibitor siRNA to stimulate both disease modifying response and long-term pain relief with a one-time intra-articular dose. IL-1RA mRNA targets are in the chondrocytes and synovium cells; Nav1.8 expressing nerves innervate into synovium and subchondral bone in OA – sites that Exo-CPC can readily target. Aim 1 will engineer cartilage targeting Exo-CPC for delivery of IL- 1RA mRNA and Nav1.8 inhibitor siRNA. Their ability to deliver IL-1RA mRNA to chondrocytes and IL-1RA protein translation efficiency will be evaluated in-vitro. Exo-CPC-Na v1.8’s ability to reduce NaV1.8 bioactivity of sensory nerves will also be evaluated. In Aim 2, their distribution intra-articular (proximity to NaV1.8-positive nerves), extra-articular, and DRG and spinal cord using partial meniscectomy NaV1.8-tdTomato reporter mice OA models will be evaluated. Additionally, their dose dependent reduction on MMP activity, neuronal excitability and pain- related behaviors, and any immunogenicity will be assessed. Aim 3 will use the determined functional doses to study the long-term disease modifying and pain-relief effects of mono and combination therapy with Exo-CPC- IL-1RA and Exo-CPC-Nav1.8 in rescuing injury induced tissue structural damage as well as in reducing pain (weight bearing asymmetry) for up to one month following IA administration in early vs. late stages (intervention at 2 vs 6 weeks) of MMT (medial meniscectomy) induced OA rats. The project paves way for utilizing the intrinsic therapeutic potential of MSC Exosomes as viral-free, non-immunogenic carriers for OA gene therapy by employing cartilage as a drug depot. Cationic exosomes can be used to deliver other OA gene targets, and can be widely used for targeting other negatively charged tissues like meniscus, ligaments, discs, fracture callus etc.

Transcriptional control of activation induced deaminase (AID) function

SUMMARY Somatic hypermutation (SHM) and class switch recombination (CSR) are vital for the generation of high affinity antibodies with appropriate effector function, protection against infection, and vaccine efficiency. They are initiated when the activation induced deaminase (AID) deaminates cytidines in single-stranded DNA in the context of transcription by RNA polymerase 2 (Pol2). Aberrant DNA deamination by AID is an important driver of genetic instability and the development of B cell malignancies. Understanding the factors and mechanisms that coordinate AID-mediated deamination with Pol2 transcription is an important objective in the study of humoral immunity and the central goal of research under this grant. Our preliminary data demonstrate that Pol2 pause factor NELF, Super Elongation Complex (SEC) components MLLT1/3, and the phosphatase module of the Integrator-protein phosphatase complex (INT-PP2A) are required for SHM, with MLLT1/3 but not NELF being required for AID binding to its chromatin targets. Our findings yield a new conceptual framework and model for AID-Pol2 collaboration in which NELF and a balance between kinase and phosphatase activities of SEC and INT-PP2A regulate Pol2 pausing/elongation to generate the critical stalled Pol2 complex on which AID acts. Further, our work has yielded major methodological advances that allow us to overcome obstacles that have stymied progress in the field. In this proposal, we take advantage of these conceptual and technical advances to pursue our central goal through the following two aims: Aim 1: Determine the molecular mechanisms by which NELF and other Pol2 regulatory factors enable AID-Pol2 collaboration and SHM/CSR. It has previously been very difficult to assess the role of cell-essential factors in SHM. By combining our new Rapid Assay for SHM (RASH) cells with degron technology, we will determine the mechanism of action of our newly discovered regulators of SHM using genomic, transcriptomic, and interaction assays that assess Pol2 distribution, phosphorylation, and activity, and the chromatin binding profiles of and interactions between AID and components of NELF, SEC, and INT-PP2A. AID and MLLT1 appear to co-associate in a complex and we will test for a direct interaction between AID and MLLT1/3. Factors will be tested for roles in CSR and validated in human cell line and germinal center B cell models and in mice. Aim 2: Hypothesis testing and deep mechanistic analysis through perturbation of the balance between Pol2 pause/arrest and elongation. We will rigorously test our new model for AID-Pol2 collaboration using degron, reconstitution, mutagenesis, and small molecular inhibitor approaches to perturb the balance between Pol2 pausing and elongation, revealing how altering NELF-Pol2 interactions and the balance between SEC kinase and INT-PP2A phosphatase activities influences SHM efficiencies and AID binding. Together, our proposed studies are significant for the development of new technologies and for understanding mechanisms of antibody gene diversification and causes of genome instability and cancer.

Validating Causality of Disputed Mitochondrial Variants in Inborn Errors of Metabolism

PROJECT SUMMARY Primary mitochondrial disease (PMD) encompasses multi-systemic disorders caused by impaired mitochondrial function. PMDs arise from pathogenic variants in either nuclear genes encoding mitochondrial proteins, or in the mitochondrial DNA (mtDNA) genome. Clinical diagnosis is challenging due to phenotypic heterogeneity, underscoring the importance of genetic diagnosis. ACMG/AMP guidelines provide a well-established framework for interpreting nuclear DNA variants while diagnosing genetic diseases. Their application to mtDNA variants, however, remains challenging due to unique features of mtDNA: maternal inheritance, heteroplasmy, threshold effects, and effect of transfer or ribosomal RNA rather than coding variants. To address these challenges, the ClinGen Mitochondrial Disease Nuclear and Mitochondrial Variant Curation Expert Panel, co-chaired by the Multi-PIs of this study, developed widely adopted ACMG/AMP revised guidelines for mtDNA variant interpretation. Over the past five years, this global expert panel has curated more than 280 mtDNA variant. Because of the lack of functional data of individual mtDNA variants in the literature, 23 previously reported pathogenic (P) variants were classified as Variants of Uncertain Significance (VUS), hindering definitive PMD diagnoses and therapeutic development. This R01 project aims to resolve the pathogenicity of these 23 mtDNA VUS through functional validation, leveraging advanced mtDNA base editing and single-cell genomics in in vitro and in vivo models. In Aim 1, we will create human 143B cell line models for 20 VUS using cutting-edge mtDNA editing techniques, optimized for efficiency and minimal off-target effects. Single-cell genomics (mtscATAC-seq and scRNA-seq) will assess heteroplasmy and genomic changes, while functional assays will evaluate mitochondrial ATP production, oxidative phosphorylation, membrane potential, and redox stress. Aim 2 will develop zebrafish models for 17 conserved VUS, characterizing phenotypic and mitochondrial outcomes to corroborate in vitro findings and PMD patient phenotypes. This study will clarify longstanding uncertainties regarding the pathogenicity of these mtDNA VUSs which were nonetheless reported to be pathogenic with often strong genetic evidence but limited functional data. The study will also establish valuable cell and zebrafish models and provide mechanistic insights of PMDs. The resulting resources will be shared with the scientific community to accelerate research and therapeutic advancements for novel precision medicine approaches for PMDs.

Investigating the role of noncoding RNAs in malaria parasites through targeted Cas13-mediated degradation

Project Summary/Abstract One of the most significant sources of morbidity and mortality throughout large regions of the developing world continues to be malaria caused by infection with mosquito-borne parasites of the genus Plasmodium. The parasite species responsible for the most severe form of the disease is P. falciparum. To avoid antibodies produced by their host and thereby maintain lengthy infections, these parasites undergo a process called antigenic variation by which they can extend an infection for over a year. This results from changes in expression of a protein called PfEMP1, the primary antigenic and virulence determinant expressed on the surface of infected red blood cells. A large, multicopy gene family called var encodes different forms of PfEMP1, and switching expression between var genes enables parasites to evade antibody recognition and destruction by the immune system. The process requires precise and coordinated regulation of transcription of each var gene, however how this is accomplished is unknown. It was recently hypothesized that a family of noncoding RNAs (ncRNAs) plays a key role in controlling the expression of each var gene and in determining the likelihood of activation of any given gene. If correct, this would represent a significant advance in our understanding of how P. falciparum controls antigenic variation and avoids immune clearance. To test this hypothesis, we propose to adapt the CRISPR/Cas13 system of targeted RNA degradation for use in P. falciparum. Similar to the extensively used CRISPR/Cas9 system, CRISPR/Cas13 employes guide RNAs to target a nuclease to a sequence-specific target, however Cas13 targets single stranded RNA rather than DNA. By applying this system to the study of var-related ncRNAs, we will degrade specific ncRNAs and determine the effect on var gene expression. Two classes of ncRNAs previously proposed to regulate var gene expression will be targeted, one called ruf6 and a second encoded by the second exon of all var genes. This will enable us to alter ncRNA expression while leaving the underlying genomic DNA untouched, thereby allowing the unambiguous attribution of any resulting phenotypes to the ncRNAs. Aim 1 will optimize the Cas13 system for P. falciparum by testing different variants of the Cas13 endonuclease for their ability to degrade mRNAs encoding fluorescent reporter proteins. We will determine both the efficiency and sequence specificity of the system. Aim 2 will apply the system to var-associated ncRNAs and quantitatively measure changes in var gene expression and transcriptional switching. If successful, the adaptation of the Cas13 system to P. falciparum will provide the malaria research community with a powerful new tool for manipulating gene expression. In addition, we will gain valuable new insights into how malaria parasites regulate var gene expression, antigenic variation and immune evasion.

Developing a novel technology for studying T cell differentiation in vivo

Summary CRISPR-based genetic screens have revolutionized our understanding of gene functions and molecular mechanisms across various biological processes. In the field of T cell biology, CRISPR screens have played a pivotal role in identifying genes that impact critical aspects, such as T cell development, differentiation, and function. However, traditional screens have struggled to distinguish genes with diverse mechanisms of action, necessitating further investigations. To address this challenge, researchers have harnessed the power of CRISPR screens combined with single-cell sequencing (scCRISPR-seq), enabling the simultaneous assessment of genetic perturbations and high-dimensional phenotypes at the single-cell level. While scCRISPR- seq has predominantly been performed in vitro using immortalized cell lines, its physiological relevance is limited due to oversimplified biological context and disparities compared to primary cells. This limitation highlights the urgent need for large-scale in vivo scCRISPR-seq with primary T cells. However, various challenges have discouraged its widespread adoption. The use of viral vectors for sgRNA delivery compromises physiological relevance, as the in vitro activation conditions fail to faithfully represent the intricate T cell priming process in vivo. Moreover, viral vector components and continuous Cas9 expression can trigger immunogenicity and cytotoxicity, leading to cell depletion and hindering long-term studies. Additionally, current scCRISPR-seq methods face technical limitations, including low editing efficiency and inadequate perturbation identity recovery rates, which impede efficient large-scale in vivo applications. Fortunately, recent advances in ribonucleoprotein complex (RNP) transfection have addressed many of these challenges. This cutting-edge technology enables efficient gene editing in primary T cells without the need for in vitro activation or permanent Cas9 expression. Leveraging the high editing efficiency of RNP transfection, the investigator’s team aims to develop a novel strategy for in vivo T cell CRISPR screens. This innovative approach involves arrayed RNP transfection and co- transfer of T cells that recognize the relevant antigens. Instead of traditional genetic barcodes, the strategy utilizes congenic markers (CD45.1/45.2 and CD90.1/CD90.2) from donor TCR transgenic T cells as "external barcodes." These markers facilitate the recovery of gene perturbation identity at the single-cell level through the application of CITE-seq. Importantly, this RNP-based strategy seamlessly integrates with existing single-cell sequencing protocols, enabling the comprehensive assessment of transcripts, epitopes, and chromatin accessibility simultaneously. To demonstrate the efficacy of this strategy, the team plans to develop two benchmarking approaches: RNP-CET-seq to investigate the role of TCR regulators in T cell exhaustion and RNP-CATE-seq to map the gene regulatory atlas of exhausted CD8 T cells. In summary, the proposed RNP- based scCRISPR-seq strategy overcomes the limitations of current approaches, enabling large-scale, multi- module in vivo genetic screens within a physiologically relevant context across various disease models.

A PROTAC Strategy to Combat Botulinum Neurotoxicity

PROJECT SUMMARY/ABSTRACT Botulinum neurotoxin (BoNT), the causative agent of botulism, is the most potent toxin known to humans. While BoNTs are widely recognized for their therapeutic and cosmetic applications, such as Botox™, their increasing use has raised concerns about iatrogenic botulism. Due to their extreme lethality, ease of production, and history of weaponization, the Centers for Disease Control and Prevention (CDC) classifies BoNTs as a Category A bioterrorism threat. Among the seven major serotypes (A-G), BoNT/A, BoNT/B, and BoNT/E account for over 95% of human botulism cases with A being the most prevalent. Despite the severity of botulism, no approved therapeutic exists to rescue intoxicated neurons. The current treatment, a heptavalent antitoxin, can only slow disease progression and requires early administration and prolonged hospitalization due to the inability of antibodies to penetrate infected cells. In the field of small- molecule inhibitors (SMIs), promising scaffolds targeting BoNT/A have been discovered, offering opportunities for further derivatization to incorporate bifunctional approaches. Developing a clinically viable therapeutic requires inhibiting the zinc (Zn2+) metalloprotease light chain (LC) as well as addressing toxin persistence. Through extensive inhibitor screening, we have identified two classes of small molecules that inhibit BoNT/A with submicromolar affinity and demonstrate efficacy in both cellular and animal models. However, the transient nature of these inhibitors necessitates the need of a sustained clearance approach. To achieve this, we propose integrating our previously identified BoNT/A LC SMIs with a targeted protein degradation (TPD) technology for toxin elimination. Based upon the background outlined, vide supra, our research strategy for the ablation of BoNT/A will be focused upon the following three specific objectives: 1) Structural Optimization – Utilize molecular docking, and structure-activity relationship (SAR) analysis to modify inhibitors for TPD ligand attachment. 2) Degrader Design – Development of ubiquitin-protease system (UPS)-based proteolysis-targeting chimeras (PROTACs) and autophagy-targeting chimeras to enhance degradation efficiency. 3) Cellular Evaluation – Assess enzyme inhibition, toxin clearance, degradation kinetics in cells.

Astrocytes: From Metabolism to Cognition

Different brain cell types exhibit distinct metabolic signatures that link energy economy to cellular function. Astrocytes and neurons, for instance, diverge dramatically in their reliance on glycolysis versus oxidative phosphorylation, underscoring that metabolic fuel efficiency is not uniform across cell types. A key factor shaping this divergence is the structural organization of the mitochondrial respiratory chain into supercomplexes. Specifically, complexes I (CI) and III (CIII) form a CI–CIII supercomplex, but the degree of this assembly varies by cell type. In neurons, CI is predominantly integrated into supercomplexes, resulting in highly efficient mitochondrial respiration and minimal reactive oxygen species (ROS) generation. Conversely, in astrocytes, a larger fraction of CI remains unassembled, freely existing apart from CIII, leading to reduced respiratory efficiency and elevated mitochondrial ROS production. Despite this apparent inefficiency, astrocytes boast a highly adaptable metabolism capable of responding to diverse stressors. Their looser CI–CIII organization allows for flexible ROS signaling, which activates antioxidant programs via transcription factors like Nrf2. This modular architecture enables astrocytes not only to balance energy production but also to support neuronal health and influence complex organismal behaviors.

Learning produces a hippocampal cognitive map in the form of an orthogonalized state machine

Cognitive maps confer animals with flexible intelligence by representing spatial, temporal, and abstract relationships that can be used to shape thought, planning, and behavior. Cognitive maps have been observed in the hippocampus, but their algorithmic form and the processes by which they are learned remain obscure. Here, we employed large-scale, longitudinal two-photon calcium imaging to record activity from thousands of neurons in the CA1 region of the hippocampus while mice learned to efficiently collect rewards from two subtly different versions of linear tracks in virtual reality. The results provide a detailed view of the formation of a cognitive map in the hippocampus. Throughout learning, both the animal behavior and hippocampal neural activity progressed through multiple intermediate stages, gradually revealing improved task representation that mirrored improved behavioral efficiency. The learning process led to progressive decorrelations in initially similar hippocampal neural activity within and across tracks, ultimately resulting in orthogonalized representations resembling a state machine capturing the inherent struture of the task. We show that a Hidden Markov Model (HMM) and a biologically plausible recurrent neural network trained using Hebbian learning can both capture core aspects of the learning dynamics and the orthogonalized representational structure in neural activity. In contrast, we show that gradient-based learning of sequence models such as Long Short-Term Memory networks (LSTMs) and Transformers do not naturally produce such orthogonalized representations. We further demonstrate that mice exhibited adaptive behavior in novel task settings, with neural activity reflecting flexible deployment of the state machine. These findings shed light on the mathematical form of cognitive maps, the learning rules that sculpt them, and the algorithms that promote adaptive behavior in animals. The work thus charts a course toward a deeper understanding of biological intelligence and offers insights toward developing more robust learning algorithms in artificial intelligence.

Dynamic endocrine modulation of the nervous system

Sex hormones are powerful neuromodulators of learning and memory. In rodents and nonhuman primates estrogen and progesterone influence the central nervous system across a range of spatiotemporal scales. Yet, their influence on the structural and functional architecture of the human brain is largely unknown. Here, I highlight findings from a series of dense-sampling neuroimaging studies from my laboratory designed to probe the dynamic interplay between the nervous and endocrine systems. Individuals underwent brain imaging and venipuncture every 12-24 hours for 30 consecutive days. These procedures were carried out under freely cycling conditions and again under a pharmacological regimen that chronically suppresses sex hormone production. First, resting state fMRI evidence suggests that transient increases in estrogen drive robust increases in functional connectivity across the brain. Time-lagged methods from dynamical systems analysis further reveals that these transient changes in estrogen enhance within-network integration (i.e. global efficiency) in several large-scale brain networks, particularly Default Mode and Dorsal Attention Networks. Next, using high-resolution hippocampal subfield imaging, we found that intrinsic hormone fluctuations and exogenous hormone manipulations can rapidly and dynamically shape medial temporal lobe morphology. Together, these findings suggest that neuroendocrine factors influence the brain over short and protracted timescales.

Relations and Predictions in Brains and Machines

Humans and animals learn and plan with flexibility and efficiency well beyond that of modern Machine Learning methods. This is hypothesized to owe in part to the ability of animals to build structured representations of their environments, and modulate these representations to rapidly adapt to new settings. In the first part of this talk, I will discuss theoretical work describing how learned representations in hippocampus enable rapid adaptation to new goals by learning predictive representations, while entorhinal cortex compresses these predictive representations with spectral methods that support smooth generalization among related states. I will also cover recent work extending this account, in which we show how the predictive model can be adapted to the probabilistic setting to describe a broader array of generalization results in humans and animals, and how entorhinal representations can be modulated to support sample generation optimized for different behavioral states. In the second part of the talk, I will overview some of the ways in which we have combined many of the same mathematical concepts with state-of-the-art deep learning methods to improve efficiency and performance in machine learning applications like physical simulation, relational reasoning, and design.

Behavioural Basis of Subjective Time Distortions

Precisely estimating event timing is essential for survival, yet temporal distortions are ubiquitous in our daily sensory experience. Here, we tested whether the relative position, duration, and distance in time of two sequentially-organized events—standard S, with constant duration, and comparison C, with duration varying trial-by-trial—are causal factors in generating temporal distortions. We found that temporal distortions emerge when the first event is shorter than the second event. Importantly, a significant interaction suggests that a longer inter-stimulus interval (ISI) helps to counteract such serial distortion effect only when the constant S is in the first position, but not if the unpredictable C is in the first position. These results imply the existence of a perceptual bias in perceiving ordered event durations, mechanistically contributing to distortion in time perception. Our results clarify the mechanisms generating time distortions by identifying a hitherto unknown duration-dependent encoding inefficiency in human serial temporal perception, something akin to a strong prior that can be overridden for highly predictable sensory events but unfolds for unpredictable ones.

Spatially-embedded recurrent neural networks reveal widespread links between structural and functional neuroscience findings

Brain networks exist within the confines of resource limitations. As a result, a brain network must overcome metabolic costs of growing and sustaining the network within its physical space, while simultaneously implementing its required information processing. To observe the effect of these processes, we introduce the spatially-embedded recurrent neural network (seRNN). seRNNs learn basic task-related inferences while existing within a 3D Euclidean space, where the communication of constituent neurons is constrained by a sparse connectome. We find that seRNNs, similar to primate cerebral cortices, naturally converge on solving inferences using modular small-world networks, in which functionally similar units spatially configure themselves to utilize an energetically-efficient mixed-selective code. As all these features emerge in unison, seRNNs reveal how many common structural and functional brain motifs are strongly intertwined and can be attributed to basic biological optimization processes. seRNNs can serve as model systems to bridge between structural and functional research communities to move neuroscientific understanding forward.

Training Dynamic Spiking Neural Network via Forward Propagation Through Time

With recent advances in learning algorithms, recurrent networks of spiking neurons are achieving performance competitive with standard recurrent neural networks. Still, these learning algorithms are limited to small networks of simple spiking neurons and modest-length temporal sequences, as they impose high memory requirements, have difficulty training complex neuron models, and are incompatible with online learning.Taking inspiration from the concept of Liquid Time-Constant (LTCs), we introduce a novel class of spiking neurons, the Liquid Time-Constant Spiking Neuron (LTC-SN), resulting in functionality similar to the gating operation in LSTMs. We integrate these neurons in SNNs that are trained with FPTT and demonstrate that thus trained LTC-SNNs outperform various SNNs trained with BPTT on long sequences while enabling online learning and drastically reducing memory complexity. We show this for several classical benchmarks that can easily be varied in sequence length, like the Add Task and the DVS-gesture benchmark. We also show how FPTT-trained LTC-SNNs can be applied to large convolutional SNNs, where we demonstrate novel state-of-the-art for online learning in SNNs on a number of standard benchmarks (S-MNIST, R-MNIST, DVS-GESTURE) and also show that large feedforward SNNs can be trained successfully in an online manner to near (Fashion-MNIST, DVS-CIFAR10) or exceeding (PS-MNIST, R-MNIST) state-of-the-art performance as obtained with offline BPTT. Finally, the training and memory efficiency of FPTT enables us to directly train SNNs in an end-to-end manner at network sizes and complexity that was previously infeasible: we demonstrate this by training in an end-to-end fashion the first deep and performant spiking neural network for object localization and recognition. Taken together, we out contribution enable for the first time training large-scale complex spiking neural network architectures online and on long temporal sequences.

Beyond Biologically Plausible Spiking Networks for Neuromorphic Computing

Biologically plausible spiking neural networks (SNNs) are an emerging architecture for deep learning tasks due to their energy efficiency when implemented on neuromorphic hardware. However, many of the biological features are at best irrelevant and at worst counterproductive when evaluated in the context of task performance and suitability for neuromorphic hardware. In this talk, I will present an alternative paradigm to design deep learning architectures with good task performance in real-world benchmarks while maintaining all the advantages of SNNs. We do this by focusing on two main features – event-based computation and activity sparsity. Starting from the performant gated recurrent unit (GRU) deep learning architecture, we modify it to make it event-based and activity-sparse. The resulting event-based GRU (EGRU) is extremely efficient for both training and inference. At the same time, it achieves performance close to conventional deep learning architectures in challenging tasks such as language modelling, gesture recognition and sequential MNIST.

Signal in the Noise: models of inter-trial and inter-subject neural variability

The ability to record large neural populations—hundreds to thousands of cells simultaneously—is a defining feature of modern systems neuroscience. Aside from improved experimental efficiency, what do these technologies fundamentally buy us? I'll argue that they provide an exciting opportunity to move beyond studying the "average" neural response. That is, by providing dense neural circuit measurements in individual subjects and moments in time, these recordings enable us to track changes across repeated behavioral trials and across experimental subjects. These two forms of variability are still poorly understood, despite their obvious importance to understanding the fidelity and flexibility of neural computations. Scientific progress on these points has been impeded by the fact that individual neurons are very noisy and unreliable. My group is investigating a number of customized statistical models to overcome this challenge. I will mention several of these models but focus particularly on a new framework for quantifying across-subject similarity in stochastic trial-by-trial neural responses. By applying this method to noisy representations in deep artificial networks and in mouse visual cortex, we reveal that the geometry of neural noise correlations is a meaningful feature of variation, which is neglected by current methods (e.g. representational similarity analysis).

General purpose event-based architectures for deep learning

Biologically plausible spiking neural networks (SNNs) are an emerging architecture for deep learning tasks due to their energy efficiency when implemented on neuromorphic hardware. However, many of the biological features are at best irrelevant and at worst counterproductive when evaluated in the context of task performance and suitability for neuromorphic hardware. In this talk, I will present an alternative paradigm to design deep learning architectures with good task performance in real-world benchmarks while maintaining all the advantages of SNNs. We do this by focusing on two main features -- event-based computation and activity sparsity. Starting from the performant gated recurrent unit (GRU) deep learning architecture, we modify it to make it event-based and activity-sparse. The resulting event-based GRU (EGRU) is extremely efficient for both training and inference. At the same time, it achieves performance close to conventional deep learning architectures in challenging tasks such as language modelling, gesture recognition and sequential MNIST

Multi-level theory of neural representations in the era of large-scale neural recordings: Task-efficiency, representation geometry, and single neuron properties

A central goal in neuroscience is to understand how orchestrated computations in the brain arise from the properties of single neurons and networks of such neurons. Answering this question requires theoretical advances that shine light into the ‘black box’ of representations in neural circuits. In this talk, we will demonstrate theoretical approaches that help describe how cognitive and behavioral task implementations emerge from the structure in neural populations and from biologically plausible neural networks. First, we will introduce an analytic theory that connects geometric structures that arise from neural responses (i.e., neural manifolds) to the neural population’s efficiency in implementing a task. In particular, this theory describes a perceptron’s capacity for linearly classifying object categories based on the underlying neural manifolds’ structural properties. Next, we will describe how such methods can, in fact, open the ‘black box’ of distributed neuronal circuits in a range of experimental neural datasets. In particular, our method overcomes the limitations of traditional dimensionality reduction techniques, as it operates directly on the high-dimensional representations, rather than relying on low-dimensionality assumptions for visualization. Furthermore, this method allows for simultaneous multi-level analysis, by measuring geometric properties in neural population data, and estimating the amount of task information embedded in the same population. These geometric frameworks are general and can be used across different brain areas and task modalities, as demonstrated in the work of ours and others, ranging from the visual cortex to parietal cortex to hippocampus, and from calcium imaging to electrophysiology to fMRI datasets. Finally, we will discuss our recent efforts to fully extend this multi-level description of neural populations, by (1) investigating how single neuron properties shape the representation geometry in early sensory areas, and by (2) understanding how task-efficient neural manifolds emerge in biologically-constrained neural networks. By extending our mathematical toolkit for analyzing representations underlying complex neuronal networks, we hope to contribute to the long-term challenge of understanding the neuronal basis of tasks and behaviors.

A model of colour appearance based on efficient coding of natural images

An object’s colour, brightness and pattern are all influenced by its surroundings, and a number of visual phenomena and “illusions” have been discovered that highlight these often dramatic effects. Explanations for these phenomena range from low-level neural mechanisms to high-level processes that incorporate contextual information or prior knowledge. Importantly, few of these phenomena can currently be accounted for when measuring an object’s perceived colour. Here we ask to what extent colour appearance is predicted by a model based on the principle of coding efficiency. The model assumes that the image is encoded by noisy spatio-chromatic filters at one octave separations, which are either circularly symmetrical or oriented. Each spatial band’s lower threshold is set by the contrast sensitivity function, and the dynamic range of the band is a fixed multiple of this threshold, above which the response saturates. Filter outputs are then reweighted to give equal power in each channel for natural images. We demonstrate that the model fits human behavioural performance in psychophysics experiments, and also primate retinal ganglion responses. Next we systematically test the model’s ability to qualitatively predict over 35 brightness and colour phenomena, with almost complete success. This implies that contrary to high-level processing explanations, much of colour appearance is potentially attributable to simple mechanisms evolved for efficient coding of natural images, and is a basis for modelling the vision of humans and other animals.

MicroRNAs as targets in the epilepsies: hits, misses and complexes

MicroRNAs are small noncoding RNAs that provide a critical layer of gene expression control. Individual microRNAs variably exert effects across networks of genes via sequence-specific binding to mRNAs, fine-tuning protein levels. This helps coordinate the timing and specification of cell fate transitions during brain development and maintains neural circuit function and plasticity by activity-dependent (re)shaping of synapses and the levels of neurotransmitter components. MicroRNA levels have been found to be altered in tissue from the epileptogenic zone resected from adults with drug-resistant focal epilepsy and this has driven efforts to explore their therapeutic potential, in particular using antisense oligonucleotide (ASOs) inhibitors termed antimirs. Here, we review the molecular mechanisms by which microRNAs control brain excitability and the latest progress towards a microRNA-based treatment for temporal lobe epilepsy. We also look at whether microRNA-based approaches could be used to treat genetic epilepsies, correcting individual genes or dysregulated pathways. Finally, we look at how cells have evolved to maximise the efficiency of the microRNA system via RNA editing, where single base changes is capable of altering the repertoire of genes under the control of a single microRNA. The findings improve our understanding of the molecular landscape of the epileptic brain and may lead to new therapies.

Efficient reuse of computations in planning

Solving complex planning problems efficiently and flexibly requires reusing expensive previous computations. The brain can do this, but how? I present a new theory that addresses this question and connects planning to hitherto distinct areas within cognitive neuroscience, such as entorhinal representation of cognitive maps and cognitive control.

A Panoramic View on Vision

Statistics of natural scenes are not uniform - their structure varies dramatically from ground to sky. It remains unknown whether these non-uniformities are reflected in the large-scale organization of the early visual system and what benefits such adaptations would confer. By deploying an efficient coding argument, we predict that changes in the structure of receptive fields across visual space increase the efficiency of sensory coding. To test this experimentally, developed a simple, novel imaging system that is indispensable for studies at this scale. In agreement with our predictions, we could show that receptive fields of retinal ganglion cells change their shape along the dorsoventral axis, with a marked surround asymmetry at the visual horizon. Our work demonstrates that, according to principles of efficient coding, the panoramic structure of natural scenes is exploited by the retina across space and cell-types.

Neural circuits for novel choices and for choice speed and accuracy changes in macaques

While most experimental tasks aim at isolating simple cognitive processes to study their neural bases, naturalistic behaviour is often complex and multidimensional. I will present two studies revealing previously uncharacterised neural circuits for decision-making in macaques. This was possible thanks to innovative experimental tasks eliciting sophisticated behaviour, bridging the human and non-human primate research traditions. Firstly, I will describe a specialised medial frontal circuit for novel choice in macaques. Traditionally, monkeys receive extensive training before neural data can be acquired, while a hallmark of human cognition is the ability to act in novel situations. I will show how this medial frontal circuit can combine the values of multiple attributes for each available novel item on-the-fly to enable efficient novel choices. This integration process is associated with a hexagonal symmetry pattern in the BOLD response, consistent with a grid-like representation of the space of all available options. We prove the causal role played by this circuit by showing that focussed transcranial ultrasound neuromodulation impairs optimal choice based on attribute integration and forces the subjects to default to a simpler heuristic decision strategy. Secondly, I will present an ongoing project addressing the neural mechanisms driving behaviour shifts during an evidence accumulation task that requires subjects to trade speed for accuracy. While perceptual decision-making in general has been thoroughly studied, both cognitively and neurally, the reasons why speed and/or accuracy are adjusted, and the associated neural mechanisms, have received little attention. We describe two orthogonal dimensions in which behaviour can vary (traditional speed-accuracy trade-off and efficiency) and we uncover independent neural circuits concerned with changes in strategy and fluctuations in the engagement level. The former involves the frontopolar cortex, while the latter is associated with the insula and a network of subcortical structures including the habenula.

Structure, Function, and Learning in Distributed Neuronal Networks

A central goal in neuroscience is to understand how orchestrated computations in the brain arise from the properties of single neurons and networks of such neurons. Answering this question requires theoretical advances that shine light into the ‘black box’ of neuronal networks. In this talk, I will demonstrate theoretical approaches that help describe how cognitive and behavioral task implementations emerge from structure in neural populations and from biologically plausible learning rules. First, I will introduce an analytic theory that connects geometric structures that arise from neural responses (i.e., neural manifolds) to the neural population’s efficiency in implementing a task. In particular, this theory describes how easy or hard it is to discriminate between object categories based on the underlying neural manifolds’ structural properties. Next, I will describe how such methods can, in fact, open the ‘black box’ of neuronal networks, by showing how we can understand a) the role of network motifs in task implementation in neural networks and b) the role of neural noise in adversarial robustness in vision and audition. Finally, I will discuss my recent efforts to develop biologically plausible learning rules for neuronal networks, inspired by recent experimental findings in synaptic plasticity. By extending our mathematical toolkit for analyzing representations and learning rules underlying complex neuronal networks, I hope to contribute toward the long-term challenge of understanding the neuronal basis of behaviors.

NMC4 Short Talk: Predictive coding is a consequence of energy efficiency in recurrent neural networks

Predictive coding represents a promising framework for understanding brain function, postulating that the brain continuously inhibits predictable sensory input, ensuring a preferential processing of surprising elements. A central aspect of this view on cortical computation is its hierarchical connectivity, involving recurrent message passing between excitatory bottom-up signals and inhibitory top-down feedback. Here we use computational modelling to demonstrate that such architectural hard-wiring is not necessary. Rather, predictive coding is shown to emerge as a consequence of energy efficiency, a fundamental requirement of neural processing. When training recurrent neural networks to minimise their energy consumption while operating in predictive environments, the networks self-organise into prediction and error units with appropriate inhibitory and excitatory interconnections and learn to inhibit predictable sensory input. We demonstrate that prediction units can reliably be identified through biases in their median preactivation, pointing towards a fundamental property of prediction units in the predictive coding framework. Moving beyond the view of purely top-down driven predictions, we demonstrate via virtual lesioning experiments that networks perform predictions on two timescales: fast lateral predictions among sensory units and slower prediction cycles that integrate evidence over time. Our results, which replicate across two separate data sets, suggest that predictive coding can be interpreted as a natural consequence of energy efficiency. More generally, they raise the question which other computational principles of brain function can be understood as a result of physical constraints posed by the brain, opening up a new area of bio-inspired, machine learning-powered neuroscience research.

NMC4 Short Talk: Rank similarity filters for computationally-efficient machine learning on high dimensional data

Real world datasets commonly contain nonlinearly separable classes, requiring nonlinear classifiers. However, these classifiers are less computationally efficient than their linear counterparts. This inefficiency wastes energy, resources and time. We were inspired by the efficiency of the brain to create a novel type of computationally efficient Artificial Neural Network (ANN) called Rank Similarity Filters. They can be used to both transform and classify nonlinearly separable datasets with many datapoints and dimensions. The weights of the filters are set using the rank orders of features in a datapoint, or optionally the 'confusion' adjusted ranks between features (determined from their distributions in the dataset). The activation strength of a filter determines its similarity to other points in the dataset, a measure based on cosine similarity. The activation of many Rank Similarity Filters transforms samples into a new nonlinear space suitable for linear classification (Rank Similarity Transform (RST)). We additionally used this method to create the nonlinear Rank Similarity Classifier (RSC), which is a fast and accurate multiclass classifier, and the nonlinear Rank Similarity Probabilistic Classifier (RSPC), which is an extension to the multilabel case. We evaluated the classifiers on multiple datasets and RSC is competitive with existing classifiers but with superior computational efficiency. Code for RST, RSC and RSPC is open source and was written in Python using the popular scikit-learn framework to make it easily accessible (https://github.com/KatharineShapcott/rank-similarity). In future extensions the algorithm can be applied to hardware suitable for the parallelization of an ANN (GPU) and a Spiking Neural Network (neuromorphic computing) with corresponding performance gains. This makes Rank Similarity Filters a promising biologically inspired solution to the problem of efficient analysis of nonlinearly separable data.

Neural network models of binocular depth perception

Our visual experience of living in a three-dimensional world is created from the information contained in the two-dimensional images projected into our eyes. The overlapping visual fields of the two eyes mean that their images are highly correlated, and that the small differences that are present represent an important cue to depth. Binocular neurons encode this information in a way that both maximises efficiency and optimises disparity tuning for the depth structures that are found in our natural environment. Neural network models provide a clear account of how these binocular neurons encode the local binocular disparity in images. These models can be expanded to multi-layer models that are sensitive to salient features of scenes, such as the orientations and discontinuities between surfaces. These deep neural network models have also shown the importance of binocular disparity for the segmentation of images into separate objects, in addition to the estimation of distance. These results demonstrate the usefulness of machine learning approaches as a tool for understanding biological vision.

Design principles of adaptable neural codes

Behavior relies on the ability of sensory systems to infer changing properties of the environment from incoming sensory stimuli. However, the demands that detecting and adjusting to changes in the environment place on a sensory system often differ from the demands associated with performing a specific behavioral task. This necessitates neural coding strategies that can dynamically balance these conflicting needs. I will discuss our ongoing theoretical work to understand how this balance can best be achieved. We connect ideas from efficient coding and Bayesian inference to ask how sensory systems should dynamically allocate limited resources when the goal is to optimally infer changing latent states of the environment, rather than reconstruct incoming stimuli. We use these ideas to explore dynamic tradeoffs between the efficiency and speed of sensory adaptation schemes, and the downstream computations that these schemes might support. Finally, we derive families of codes that balance these competing objectives, and we demonstrate their close match to experimentally-observed neural dynamics during sensory adaptation. These results provide a unifying perspective on adaptive neural dynamics across a range of sensory systems, environments, and sensory tasks.

Efficient GPU training of SNNs using approximate RTRL

Last year’s SNUFA workshop report concluded “Moving toward neuron numbers comparable with biology and applying these networks to real-world data-sets will require the development of novel algorithms, software libraries, and dedicated hardware accelerators that perform well with the specifics of spiking neural networks” [1]. Taking inspiration from machine learning libraries — where techniques such as parallel batch training minimise latency and maximise GPU occupancy — as well as our previous research on efficiently simulating SNNs on GPUs for computational neuroscience [2,3], we are extending our GeNN SNN simulator to pursue this vision. To explore GeNN’s potential, we use the eProp learning rule [4] — which approximates RTRL — to train SNN classifiers on the Spiking Heidelberg Digits and the Spiking Sequential MNIST datasets. We find that the performance of these classifiers is comparable to those trained using BPTT [5] and verify that the theoretical advantages of neuron models with adaptation dynamics [5] translate to improved classification performance. We then measured execution times and found that training an SNN classifier using GeNN and eProp becomes faster than SpyTorch and BPTT after less than 685 timesteps and much larger models can be trained on the same GPU when using GeNN. Furthermore, we demonstrate that our implementation of parallel batch training improves training performance by over 4⨉ and enables near-perfect scaling across multiple GPUs. Finally, we show that performing inference using a recurrent SNN using GeNN uses less energy and has lower latency than a comparable LSTM simulated with TensorFlow [6].

The Challenge and Opportunities of Mapping Cortical Layer Activity and Connectivity with fMRI

In this talk I outline the technical challenges and current solutions to layer fMRI. Specifically, I describe our acquisition strategies for maximizing resolution, spatial coverage, time efficiency as well as, perhaps most importantly, vascular specificity. Novel applications from our group, including mapping feedforward and feedback connections to M1 during task and sensory input modulation and S1 during a sensory prediction task are be shown. Layer specific activity in dorsal lateral prefrontal cortex during a working memory task is also demonstrated. Additionally, I’ll show preliminary work on mapping whole brain layer-specific resting state connectivity and hierarchy.

Fragility of the human connectome across the lifespan

The human brain network architecture can reveal crucial aspects of brain function and dysfunction. The topology of this network (known as the connectome) is shaped by a trade-off between wiring cost and network efficiency, and it has highly connected hub regions playing a prominent role in many brain disorders. By studying a landscape of plausible brain networks that preserve the wiring cost, fragile and resilient hubs can be identified. In this webinar, Dr Leonardo Gollo and Dr James Pang from Monash University will discuss this approach across the lifespan and some of its implications for neurodevelopmental and neurodegenerative diseases. Dr Leonardo Gollo is a Senior Research Fellow at the Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University. He holds an ARC Future Fellowship and his research interests include brain modelling, systems neuroscience, and connectomics. Dr James Pang is a Research Fellow at the Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University. His research interests are on combining neuroimaging and biophysical modelling to better understand the mechanisms of brain function in health and disease.

Design principles of adaptable neural codes

Behavior relies on the ability of sensory systems to infer changing properties of the environment from incoming sensory stimuli. However, the demands that detecting and adjusting to changes in the environment place on a sensory system often differ from the demands associated with performing a specific behavioral task. This necessitates neural coding strategies that can dynamically balance these conflicting needs. I will discuss our ongoing theoretical work to understand how this balance can best be achieved. We connect ideas from efficient coding and Bayesian inference to ask how sensory systems should dynamically allocate limited resources when the goal is to optimally infer changing latent states of the environment, rather than reconstruct incoming stimuli. We use these ideas to explore dynamic tradeoffs between the efficiency and speed of sensory adaptation schemes, and the downstream computations that these schemes might support. Finally, we derive families of codes that balance these competing objectives, and we demonstrate their close match to experimentally-observed neural dynamics during sensory adaptation. These results provide a unifying perspective on adaptive neural dynamics across a range of sensory systems, environments, and sensory tasks.

Receptor Costs Determine Retinal Design

Our group is interested in discovering design principles that govern the structure and function of neurons and neural circuits. We record from well-defined neurons, mainly in flies’ visual systems, to measure the molecular and cellular factors that determine relevant measures of performance, such as representational capacity, dynamic range and accuracy. We combine this empirical approach with modelling to see how the basic elements of neural systems (ion channels, second messengers systems, membranes, synapses, neurons, circuits and codes) combine to determine performance. We are investigating four general problems. How are circuits designed to integrate information efficiently? How do sensory adaptation and synaptic plasticity contribute to efficiency? How do the sizes of neurons and networks relate to energy consumption and representational capacity? To what extent have energy costs shaped neurons, sense organs and brain regions during evolution?

A function approximation perspective on neural representations

Activity patterns of neural populations in natural and artificial neural networks constitute representations of data. The nature of these representations and how they are learned are key questions in neuroscience and deep learning. In his talk, I will describe my group's efforts in building a theory of representations as feature maps leading to sample efficient function approximation. Kernel methods are at the heart of these developments. I will present applications to deep learning and neuronal data.

On climate change, multi-agent systems and the behaviour of networked control

Multi-agent reinforcement learning (MARL) has recently shown great promise as an approach to networked system control. Arguably, one of the most difficult and important tasks for which large scale networked system control is applicable is common-pool resource (CPR) management. Crucial CPRs include arable land, fresh water, wetlands, wildlife, fish stock, forests and the atmosphere, of which proper management is related to some of society’s greatest challenges such as food security, inequality and climate change. This talk will consist of three parts. In the first, we will briefly look at climate change and how it poses a significant threat to life on our planet. In the second, we will consider the potential of multi-agent systems for climate change mitigation and adaptation. And finally, in the third, we will discuss recent research from InstaDeep into better understanding the behaviour of networked MARL systems used for CPR management. More specifically, we will see how the tools from empirical game-theoretic analysis may be harnessed to analyse the differences in networked MARL systems. The results give new insights into the consequences associated with certain design choices and provide an additional dimension of comparison between systems beyond efficiency, robustness, scalability and mean control performance.

What can we further learn from the brain for artificial intelligence?

Deep learning is a prime example of how brain-inspired computing can benefit development of artificial intelligence. But what else can we learn from the brain for bringing AI and robotics to the next level? Energy efficiency and data efficiency are the major features of the brain and human cognition that today’s deep learning has yet to deliver. The brain can be seen as a multi-agent system of heterogeneous learners using different representations and algorithms. The flexible use of reactive, model-free control and model-based “mental simulation” appears to be the basis for computational and data efficiency of the brain. How the brain efficiently acquires and flexibly combines prediction and control modules is a major open problem in neuroscience and its solution should help developments of more flexible and autonomous AI and robotics.

Fast and deep neuromorphic learning with time-to-first-spike coding

Engineered pattern-recognition systems strive for short time-to-solution and low energy-to-solution characteristics. This represents one of the main driving forces behind the development of neuromorphic devices. For both them and their biological archetypes, this corresponds to using as few spikes as early as possible. The concept of few and early spikes is used as the founding principle in the time-to-first-spike coding scheme. Within this framework, we have developed a spike-timing-based learning algorithm, which we used to train neuronal networks on the mixed-signal neuromorphic platform BrainScaleS-2. We derive, from first principles, error-backpropagation-based learning in networks of leaky integrate-and-fire (LIF) neurons relying only on spike times, for specific configurations of neuronal and synaptic time constants. We explicitly examine applicability to neuromorphic substrates by studying the effects of reduced weight precision and range, as well as of parameter noise. We demonstrate the feasibility of our approach on continuous and discrete data spaces, both in software simulations and on BrainScaleS-2. This narrows the gap between previous models of first-spike-time learning and biological neuronal dynamics and paves the way for fast and energy-efficient neuromorphic applications.

E-prop: A biologically inspired paradigm for learning in recurrent networks of spiking neurons

Transformative advances in deep learning, such as deep reinforcement learning, usually rely on gradient-based learning methods such as backpropagation through time (BPTT) as a core learning algorithm. However, BPTT is not argued to be biologically plausible, since it requires to a propagate gradients backwards in time and across neurons. Here, we propose e-prop, a novel gradient-based learning method with local and online weight update rules for recurrent neural networks, and in particular recurrent spiking neural networks (RSNNs). As a result, e-prop has the potential to provide a substantial fraction of the power of deep learning to RSNNs. In this presentation, we will motivate e-prop from the perspective of recent insights in neuroscience and show how these have to be combined to form an algorithm for online gradient descent. The mathematical results will be supported by empirical evidence in supervised and reinforcement learning tasks. We will also discuss how limitations that are inherited from gradient-based learning methods, such as sample-efficiency, can be addressed by considering an evolution-like optimization that enhances learning on particular task families. The emerging learning architecture can be used to learn tasks by a single demonstration, hence enabling one-shot learning.

On temporal coding in spiking neural networks with alpha synaptic function

The timing of individual neuronal spikes is essential for biological brains to make fast responses to sensory stimuli. However, conventional artificial neural networks lack the intrinsic temporal coding ability present in biological networks. We propose a spiking neural network model that encodes information in the relative timing of individual neuron spikes. In classification tasks, the output of the network is indicated by the first neuron to spike in the output layer. This temporal coding scheme allows the supervised training of the network with backpropagation, using locally exact derivatives of the postsynaptic spike times with respect to presynaptic spike times. The network operates using a biologically-plausible alpha synaptic transfer function. Additionally, we use trainable synchronisation pulses that provide bias, add flexibility during training and exploit the decay part of the alpha function. We show that such networks can be trained successfully on noisy Boolean logic tasks and on the MNIST dataset encoded in time. The results show that the spiking neural network outperforms comparable spiking models on MNIST and achieves similar quality to fully connected conventional networks with the same architecture. We also find that the spiking network spontaneously discovers two operating regimes, mirroring the accuracy-speed trade-off observed in human decision-making: a slow regime, where a decision is taken after all hidden neurons have spiked and the accuracy is very high, and a fast regime, where a decision is taken very fast but the accuracy is lower. These results demonstrate the computational power of spiking networks with biological characteristics that encode information in the timing of individual neurons. By studying temporal coding in spiking networks, we aim to create building blocks towards energy-efficient and more complex biologically-inspired neural architectures.

Effective and Efficient Computation with Multiple-timescale Spiking Recurrent Neural Networks

The emergence of brain-inspired neuromorphic computing as a paradigm for edge AI is motivating the search for high-performance and efficient spiking neural networks to run on this hardware. However, compared to classical neural networks in deep learning, current spiking neural networks lack competitive performance in compelling areas. Here, for sequential and streaming tasks, we demonstrate how spiking recurrent neural networks (SRNN) using adaptive spiking neurons are able to achieve state-of-the-art performance compared to other spiking neural networks and almost reach or exceed the performance of classical recurrent neural networks (RNNs) while exhibiting sparse activity. From this, we calculate a 100x energy improvement for our SRNNs over classical RNNs on the harder tasks. We find in particular that adapting the timescales of spiking neurons is crucial for achieving such performance, and we demonstrate the performance for SRNNs for different spiking neuron models.

Neuronal morphology imposes a tradeoff between stability, accuracy and efficiency of synaptic scaling

Synaptic scaling is a homeostatic normalization mechanism that preserves relative synaptic strengths by adjusting them with a common factor. This multiplicative change is believed to be critical, since synaptic strengths are involved in learning and memory retention. Further, this homeostatic process is thought to be crucial for neuronal stability, playing a stabilizing role in otherwise runaway Hebbian plasticity [1-3]. Synaptic scaling requires a mechanism to sense total neuron activity and globally adjust synapses to achieve some activity set-point [4]. This process is relatively slow, which places limits on its ability to stabilize network activity [5]. Here we show that this slow response is inevitable in realistic neuronal morphologies. Furthermore, we reveal that global scaling can in fact be a source of instability unless responsiveness or scaling accuracy are sacrificed." "A neuron with tens of thousands of synapses must regulate its own excitability to compensate for changes in input. The time requirement for global feedback can introduce critical phase lags in a neuron’s response to perturbation. The severity of phase lag increases with neuron size. Further, a more expansive morphology worsens cell responsiveness and scaling accuracy, especially in distal regions of the neuron. Local pools of reserve receptors improve efficiency, potentiation, and scaling, but this comes at a cost. Trafficking large quantities of receptors requires time, exacerbating the phase lag and instability. Local homeostatic feedback mitigates instability, but this too comes at the cost of reducing scaling accuracy." "Realization of the phase lag instability requires a unified model of synaptic scaling, regulation, and transport. We present such a model with global and local feedback in realistic neuron morphologies (Fig. 1). This combined model shows that neurons face a tradeoff between stability, accuracy, and efficiency. Global feedback is required for synaptic scaling but favors either system stability or efficiency. Large receptor pools improve scaling accuracy in large morphologies but worsen both stability and efficiency. Local feedback improves the stability-efficiency tradeoff at the cost of scaling accuracy. This project introduces unexplored constraints on neuron size, morphology, and synaptic scaling that are weakened by an interplay between global and local feedback.

The butterfly strikes back: neurons doing 'network' computation

We live in the age of the network: Internet social neural ecosystems. This has become one of the main metaphors for how we think about complex systems. This view also dominates the account of brain function. The role of neuronsdescribed by Cajal as the "butterflies of the soul" has become diminished to leaky integrate-and-fire point objects in many models of neural network computation. It is perhaps not surprising that networkexplanations of neural phenomena use neurons as elementary particles andascribe all their wonderful capabilities to their interactions in a network. In the network view the Connectome defines the brain and the butterflies have no role. In this talk I'd like to reclaim some key computations from the networkand return them to their rightful place at the cellular and subcellular level. I'll start with a provocative look at potential computational capacity ofdifferent kinds of brain computation: network vs. subcellular. I'll then consider different levels of pattern and sequence computationwith a glimpse of the efficiency of the subcellular solutions. Finally I propose that there is a suggestive mapping between entire nodesof deep networks to individual neurons. This in my view is how we can walk around with 1.3 litres and 20 watts of installed computational capacity still doing far more than giant AI server farms.

Inferring Brain Rhythm Circuitry and Burstiness

Bursts in gamma and other frequency ranges are thought to contribute to the efficiency of working memory or communication tasks. Abnormalities in bursts have also been associated with motor and psychiatric disorders. The determinants of burst generation are not known, specifically how single cell and connectivity parameters influence burst statistics and the corresponding brain states. We first present a generic mathematical model for burst generation in an excitatory-inhibitory (EI) network with self-couplings. The resulting equations for the stochastic phase and envelope of the rhythm’s fluctuations are shown to depend on only two meta-parameters that combine all the network parameters. They allow us to identify different regimes of amplitude excursions, and to highlight the supportive role that network finite-size effects and noisy inputs to the EI network can have. We discuss how burst attributes, such as their durations and peak frequency content, depend on the network parameters. In practice, the problem above follows the a priori challenge of fitting such E-I spiking networks to single neuron or population data. Thus, the second part of the talk will discuss a novel method to fit mesoscale dynamics using single neuron data along with a low-dimensional, and hence statistically tractable, single neuron model. The mesoscopic representation is obtained by approximating a population of neurons as multiple homogeneous ‘pools’ of neurons, and modelling the dynamics of the aggregate population activity within each pool. We derive the likelihood of both single-neuron and connectivity parameters given this activity, which can then be used to either optimize parameters by gradient ascent on the log-likelihood, or to perform Bayesian inference using Markov Chain Monte Carlo (MCMC) sampling. We illustrate this approach using an E-I network of generalized integrate-and-fire neurons for which mesoscopic dynamics have been previously derived. We show that both single-neuron and connectivity parameters can be adequately recovered from simulated data.

Computational modelling of dentate granule cells reveals Pareto optimal trade-off between pattern separation and energy efficiency (economy)

Bernstein Conference 2024

Balancing safety and efficiency in human decision-making.

COSYNE 2022

Neuromodulatory changes in the efficiency of information transmission at visual synapses

COSYNE 2022

Neuromodulatory changes in the efficiency of information transmission at visual synapses

COSYNE 2022

Adaptive coding efficiency through joint gain control in neural populations

COSYNE 2023

On the benefits of analog spikes: an information efficiency perspective

COSYNE 2025

The combined efficiency of Thymoquinone and Curcumin on Glioblastoma cells

Effects of sustained cognitive activity on executive functions in healthy aged adults : toward a gain of 10 years of cognitive efficiency

Filtering Efficiency as an Underlying Mechanism of Model-based Working Memory Training in Healthy Old Adults

High-efficiency transdifferentiation of Human Dental Pulp Stem Cells into functional motor neuron via Optogenetics Stimulation

Odor Background Increases the Pheromone Coding Efficiency in Moth Olfactory Neurons

Rapid, high-efficiency differentiation of motor neurons from human pluripotent stem cells

Relationships between the efficiency of temporal information processing on milli- and supra-second time domains

Resting EEG theta activity as an electrophysiological marker of individual differences in efficiency of Temporal Information Processing

Advanced metamodelling on the o2S2PARC computational neurosciences platform facilitates stimulation selectivity and power efficiency optimization and intelligent control

FENS Forum 2024

Comparison of modulation efficiency between normal and degenerated primate retina

FENS Forum 2024

Development of a next-generation bidirectional neurobiohybrid interface with optimized energy efficiency enabling real-time adaptive neuromodulation

FENS Forum 2024

DJ-1-mediated metabolic efficiency determined the vulnerability of midbrain dopaminergic neurons in Parkinson’s disease

FENS Forum 2024

Improving perceptual learning efficiency with brief memory reactivations engages distinct neural mechanisms

FENS Forum 2024

Investigating the efficiency of a mitochondria booster to improve anxiety-related behaviors: Accumbal metabolic and neurobiological mechanisms

FENS Forum 2024

Reduced routing efficiency in the right fronto-parietal attentional network during distractor suppression in mild cognitive impairment

FENS Forum 2024

Pruning for efficiency in Hopfield networks

Neuromatch 5