Neural circuit mechanisms of hierarchical sequence learning tested on large-scale recording data

Asabuki, Toshitake; Kokate, Prajakta; Fukai, Tomoki

doi:10.1371/journal.pcbi.1010214

Cited by 12 publications

(6 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is desirable e.g. in spiking models that employ structural (Deger et al, 2012; Gallinaro et al, 2022) or synaptic (Vogels et al, 2011; Zenke and Ganguli, 2018; Sacramento et al, 2018; Asabuki et al, 2022; Illing et al, 2021) plasticity to support continual learning and the formation and recall of short-term (on the time scale of minutes), middle-term (hours) or long-term (days) associative memories. Similarly, simulating nervous system control of behaving agents in approaches to computational neuroethology may require biological model time scales of minutes to hours or days.…”

Section: Discussionmentioning

confidence: 99%

Efficient parameter calibration and real-time simulation of large scale spiking neural networks with GeNN and NEST

Schmitt

Rostami

Nawrot

2022

Preprint

View full text Add to dashboard Cite

Spiking neural networks (SNN) represent the state-of-the-art approach to the biologically realistic modeling of nervous system function. The systematic calibration for multiple free model parameters to achieve robust network function demands high computing power and large memory resources. Special requirements arise from closed-loop model simulation in virtual environments, and from real-time simulation in robotic application. Here, we compare two complementary approaches to efficient large scale and real-time SNN simulation. The widely used NEural Simulation Technology (NEST) parallelizes simulation across multiple CPU cores. The GPU-enhanced Neural Network (GeNN) simulator uses the highlyparallel GPU-based architecture to gain simulation speed. We quantify fixed and variable simulation costs on different hardware configurations. As benchmark model we us a spiking cortical attractor network with a topology of densely connected excitatory and inhibitory neuron clusters. We show that simulation time scales linear with the simulated biological time and, for large networks, approximately linear with the model size as dominated by the number of synaptic connections. Additional fixed costs with GeNN for model compilation are independent of model size while fixed costs with NEST for model building increase linear with model size. We demonstrate how GeNN can be used for simulating networks with up to 3, 5 million neurons (> 3 × 1012 synapses) on a high-end GPU, and up to 250 thousand neurons (25 billion synapses) on a low-cost GPU. Real-time simulation was achieved for networks of more than 90 k neurons. We conclude that GeNN is in advantage for large networks and real-time applications while NEST plays out its strengths in a high degree of flexibility ideal for prototyping, and easy access for non-expert programmers.

show abstract

Section: Discussionmentioning

confidence: 99%

Efficient parameter calibration and real-time simulation of large scale spiking neural networks with GeNN and NEST

Schmitt

Rostami

Nawrot

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Simulations were performed using MOOSE. For Figure 5B, inputs were delivered either as an ordered pattern [0, 1, 2, 3, 4], or a scrambled pattern [4, 1, 0, 3, 2] with consecutive inputs spaced at a distance of 3 μm on a 100 μm long dendrite that had a diameter of 10 μm. The time interval between inputs was 2s.…”

Section: Methodsmentioning

confidence: 99%

“…Neurons receiving sequential connectivity for the ordered stimulus [1,2,3,4] (PSCSD neurons) showed slightly higher activations on average relative to neurons that did not receive such connectivity (PSCSD C neurons) (Figure 5A,D,E). PSCSD neurons also showed higher activation to sequential stimuli than to scrambled stimuli, e.g., [3,1,2,4] (Figure 5B, G, I) or reverse stimuli [4,3,2,1] (Figure 5 C, H, I). PSCSD neurons showed slightly higher selectivity for the ordered sequence than PSCSD C neurons (Figure 5J).…”

Section: Sequential Convergence Of Inputsmentioning

confidence: 98%

“…While connectivity is a necessary substrate for clustered dendritic computations, we show that factors such as background activity, length and timescales of postsynaptic mechanisms, and dendritic nonlinearities can play a key role in enabling the network to tap into such computations. In addition, our analysis can guide the design of biologically-inspired networks of neurons with dendrites for solving a variety of interesting problems (Chavlis and Poirazi 2024;Iyer et al 2022;Asabuki and Fukai 2020;Asabuki, Kokate, and Fukai 2022).…”

Section: Testing Connectivity Predictions Using Structural Connectivitymentioning

confidence: 99%

See 1 more Smart Citation

Discriminating neural ensemble patterns through dendritic computations in randomly connected feedforward networks

Somashekar,

Bhalla

2024

Preprint

View full text Add to dashboard Cite

Co-active or temporally ordered neural ensembles are a signature of salient sensory, motor, and cognitive events. Local convergence of such patterned activity as synaptic clusters on dendrites could help single neurons harness the potential of dendritic nonlinearities to decode neural activity patterns. We combined theory and simulations to assess the likelihood of whether projections from neural ensembles could converge onto synaptic clusters even in networks with random connectivity. Using rat hippocampal and cortical network statistics, we show that clustered convergence of axons from 3-4 different co-active ensembles is likely even in randomly connected networks, leading to representation of arbitrary input combinations in at least ten target neurons in a 100,000 population. In the presence of larger ensembles, spatiotemporally ordered convergence of 3-5 axons from temporally ordered ensembles is also likely. These active clusters result in higher neuronal activation in the presence of strong dendritic nonlinearities and low background activity. We mathematically and computationally demonstrate a tight interplay between network connectivity, spatiotemporal scales of subcellular electrical and chemical mechanisms, dendritic nonlinearities, and uncorrelated background activity. We suggest that dendritic clustered and sequence computation is pervasive, but its expression as somatic selectivity requires confluence of physiology, background activity, and connectomics.

show abstract

“…Our results show that GPU-based simulation can support the efficient simulation over long biological model times (Figure 3). This is desirable, e.g., in spiking models that employ structural (Deger et al, 2012;Gallinaro et al, 2022) or synaptic (Vogels et al, 2011;Sacramento et al, 2018;Zenke and Ganguli, 2018;Illing et al, 2021;Asabuki et al, 2022) plasticity to support continual learning and the formation and recall of short-term (on the time scale of minutes), middle-term (hours), or long-term (days) associative memories. Similarly, simulating nervous system control of behaving agents in approaches to computational neuroethology may require biological model time scales of minutes to hours or days.…”

Section: E Cient Long-duration and Real-time Simulation On The Gpumentioning

confidence: 99%

Efficient parameter calibration and real-time simulation of large-scale spiking neural networks with GeNN and NEST

Schmitt

Rostami²,

Nawrot³

2023

Front. Neuroinform.

View full text Add to dashboard Cite

Spiking neural networks (SNNs) represent the state-of-the-art approach to the biologically realistic modeling of nervous system function. The systematic calibration for multiple free model parameters is necessary to achieve robust network function and demands high computing power and large memory resources. Special requirements arise from closed-loop model simulation in virtual environments and from real-time simulation in robotic application. Here, we compare two complementary approaches to efficient large-scale and real-time SNN simulation. The widely used NEural Simulation Tool (NEST) parallelizes simulation across multiple CPU cores. The GPU-enhanced Neural Network (GeNN) simulator uses the highly parallel GPU-based architecture to gain simulation speed. We quantify fixed and variable simulation costs on single machines with different hardware configurations. As a benchmark model, we use a spiking cortical attractor network with a topology of densely connected excitatory and inhibitory neuron clusters with homogeneous or distributed synaptic time constants and in comparison to the random balanced network. We show that simulation time scales linearly with the simulated biological model time and, for large networks, approximately linearly with the model size as dominated by the number of synaptic connections. Additional fixed costs with GeNN are almost independent of model size, while fixed costs with NEST increase linearly with model size. We demonstrate how GeNN can be used for simulating networks with up to 3.5 · 106 neurons (> 3 · 1012synapses) on a high-end GPU, and up to 250, 000 neurons (25 · 109 synapses) on a low-cost GPU. Real-time simulation was achieved for networks with 100, 000 neurons. Network calibration and parameter grid search can be efficiently achieved using batch processing. We discuss the advantages and disadvantages of both approaches for different use cases.

show abstract

Neural circuit mechanisms of hierarchical sequence learning tested on large-scale recording data

Cited by 12 publications

References 50 publications

Efficient parameter calibration and real-time simulation of large scale spiking neural networks with GeNN and NEST

Efficient parameter calibration and real-time simulation of large scale spiking neural networks with GeNN and NEST

Discriminating neural ensemble patterns through dendritic computations in randomly connected feedforward networks

Efficient parameter calibration and real-time simulation of large-scale spiking neural networks with GeNN and NEST

Contact Info

Product

Resources

About