Encoding symbolic sequences with spiking neural reservoirs

Front. Comput. Neurosci.

et al. 2019

Self Cite

Neurobiological systems rely on hierarchical and modular architectures to carry out intricate computations using minimal resources. A prerequisite for such systems to operate adequately is the capability to reliably and efficiently transfer information across multiple modules. Here, we study the features enabling a robust transfer of stimulus representations in modular networks of spiking neurons, tuned to operate in a balanced regime. To capitalize on the complex, transient dynamics that such networks exhibit during active processing, we apply reservoir computing principles and probe the systems' computational efficacy with specific tasks. Focusing on the comparison of random feed-forward connectivity and biologically inspired topographic maps, we find that, in a sequential set-up, structured projections between the modules are strictly necessary for information to propagate accurately to deeper modules. Such mappings not only improve computational performance and efficiency, they also reduce response variability, increase robustness against interference effects, and boost memory capacity. We further investigate how information from two separate input streams is integrated and demonstrate that it is more advantageous to perform non-linear computations on the input locally, within a given module, and subsequently transfer the result downstream, rather than transferring intermediate information and performing the computation downstream. Depending on how information is integrated early on in the system, the networks achieve similar task-performance using different strategies, indicating that the dimensionality of the neural responses does not necessarily correlate with nonlinear integration, as predicted by previous studies. These findings highlight a key role of topographic maps in supporting fast, robust, and accurate neural communication over longer distances. Given the prevalence of such structural feature, particularly in the sensory systems, elucidating their functional purpose remains an important challenge toward which this work provides relevant, new insights. At the same time, these results shed new light on important requirements for designing functional hierarchical spiking networks.

Section: Stimulus Input and Computational Tasksmentioning

confidence: 99%

Passing the Message: Representation Transfer in Modular Balanced Networks

Zajzon

Mahmoudian

Front. Comput. Neurosci.

et al. 2019

Self Cite

“…Relatively low-dimensional input streams are thus non-linearly projected onto the circuit's high-dimensional representational space. Through this expansive transformation, the neural substrate can develop suitable dynamic representations (Duarte and Morrison, 2014 ; Duarte et al, 2018 ) and resolve non-linearities such that classes that are not linearly separable in the input space can be separated in the system's representational space. This property relies on the characteristics of the neural substrate, acting as a non-linear operator, and the ensuing input-driven dynamics (Maass et al, 2002 ).…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Learning and Clustered Connectivity Enhance Reinforcement Learning in Spiking Neural Networks

Weidel

Duarte

Front. Comput. Neurosci.

2021

Self Cite

Reinforcement learning is a paradigm that can account for how organisms learn to adapt their behavior in complex environments with sparse rewards. To partition an environment into discrete states, implementations in spiking neuronal networks typically rely on input architectures involving place cells or receptive fields specified ad hoc by the researcher. This is problematic as a model for how an organism can learn appropriate behavioral sequences in unknown environments, as it fails to account for the unsupervised and self-organized nature of the required representations. Additionally, this approach presupposes knowledge on the part of the researcher on how the environment should be partitioned and represented and scales poorly with the size or complexity of the environment. To address these issues and gain insights into how the brain generates its own task-relevant mappings, we propose a learning architecture that combines unsupervised learning on the input projections with biologically motivated clustered connectivity within the representation layer. This combination allows input features to be mapped to clusters; thus the network self-organizes to produce clearly distinguishable activity patterns that can serve as the basis for reinforcement learning on the output projections. On the basis of the MNIST and Mountain Car tasks, we show that our proposed model performs better than either a comparable unclustered network or a clustered network with static input projections. We conclude that the combination of unsupervised learning and clustered connectivity provides a generic representational substrate suitable for further computation.

“…Relatively low-dimensional input streams are thus non-linearly projected onto the circuit’s high-dimensional representational space. Through this expansive transformation, the neural substrate can develop suitable dynamic representations (Duarte & Morrison, 2014; Duarte et al, 2018) and resolve non-linearities such that classes that are not linearly separable in the input space can be separated in the system’s representational space. This property, commonly referred to as the kernel trick , relies on the characteristics of the neural substrate, acting as a non-linear operator, and the ensuing input-driven dynamics (Maass et al, 2002).…”

Section: Introductionmentioning

confidence: 99%

Unsupervised learning and clustered connectivity enhance reinforcement learning in spiking neural networks

Weidel

Duarte

2020

Preprint

Self Cite

2Reinforcement learning is a learning paradigm that can account for how organisms learn to 3 adapt their behavior in complex environments with sparse rewards. However, implementations in 4 spiking neuronal networks typically rely on input architectures involving place cells or receptive 5 fields. This is problematic, as such approaches either scale badly as the environment grows in size 6 or complexity, or presuppose knowledge on how the environment should be partitioned. Here, we 7 propose a learning architecture that combines unsupervised learning on the input projections with 8 clustered connectivity within the representation layer. This combination allows input features to 9 be mapped to clusters; thus the network self-organizes to produce task-relevant activity patterns 10 that can serve as the basis for reinforcement learning on the output projections. On the basis 11 of the MNIST and Mountain Car tasks, we show that our proposed model performs better than 12 either a comparable unclustered network or a clustered network with static input projections. We 13 conclude that the combination of unsupervised learning and clustered connectivity provides a 14 generic representational substrate suitable for further computation. 15