Cognitive swarming: an approach from the theoretical neuroscience of hippocampal function

Monaco, Joseph D.; Hwang, Grace M.; Schultz, Kevin; Zhang, Kechen

doi:10.1117/12.2518966

Cited by 10 publications

(7 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…equation ( 12)) and could be used to drive spike generation, if desired, in future models. Phase-based organization additionally leveraged the expressive complexity of mobile oscillators revealed by the swarmalator formalism (O'Keeffe et al, 2017;Monaco et al, 2019b). The separation of position vs. self-localization allowed swarm motion dynamics to be interpreted as Hebbian learning in an oscillatory place-coding neural network (Section 2.2).…”

Section: Discussionmentioning

confidence: 99%

Cognitive swarming in complex environments with attractor dynamics and oscillatory computing

et al. 2020

Self Cite

View full text Add to dashboard Cite

Neurobiological theories of spatial cognition developed with respect to recording data from relatively small and/or simplistic environments compared to animals' natural habitats. It has been unclear how to extend theoretical models to large or complex spaces. Complementarily, in autonomous systems technology, applications have been growing for distributed control methods that scale to large numbers of lowfootprint mobile platforms. Animals and many-robot groups must solve common problems of navigating complex and uncertain environments. Here, we introduce the 'NeuroSwarms' control framework to investigate whether adaptive, autonomous swarm control of minimal artificial agents can be achieved by direct analogy to neural circuits of rodent spatial cognition. NeuroSwarms analogizes agents to neurons and swarming groups to recurrent networks. We implemented neuron-like agent interactions in which mutually visible agents operate as if they were reciprocally-connected place cells in an attractor network. We attributed a phase state to agents to enable patterns of oscillatory synchronization similar to hippocampal models of theta-rhythmic (5-12 Hz) sequence generation. We demonstrate that multi-agent swarming and rewardapproach dynamics can be expressed as a mobile form of Hebbian learning and that NeuroSwarms supports a single-entity paradigm that directly informs theoretical models of animal cognition. We present emergent behaviors including phase-organized rings and trajectory sequences that interact with environmental cues and geometry in large, fragmented mazes. Thus, Neu-roSwarms is a model artificial spatial system that integrates autonomous control and theoretical neuroscience to potentially uncover common principles to advance both domains.

show abstract

Section: Discussionmentioning

confidence: 99%

Cognitive swarming in complex environments with attractor dynamics and oscillatory computing

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Swarmalators were originally defined as hypothetical entities that move in space and have an internal oscillatory degree of freedom, being able to swarm and sync 15 . More recently, variations of this model have been proposed to describe cognitive systems 19 and as an implementation of artificial systems [20][21][22] . The nature of these applications reveals the importance of understanding their robustness, fidelity and response to external inputs.…”

Section: Discussionmentioning

confidence: 99%

Synchronization and spatial patterns in forced swarmalators

Lizárraga

Aguiar

2020

Chaos: An Interdisciplinary Journal of Nonlinear Science

View full text Add to dashboard Cite

Swarmalators are particles that exhibit coordinated motion and, at the same time, synchronize their intrinsic behavior, represented by internal phases. Here, we study the effects produced by an external periodic stimulus over a system of swarmalators that move in two dimensions. The system represents, for example, a swarm of fireflies in the presence of an external light source that flashes at a fixed frequency. If the spatial movement is ignored, the dynamics of the internal variables are equivalent to those of Kuramoto oscillators. In this case, the phases tend to synchronize and lock to the external stimulus if its intensity is sufficiently large. Here, we show that in a system of swarmalators, the force also shifts the phases and angular velocities leading to synchronization with the external frequency. However, the correlation between phase and spatial location decreases with the intensity of the force, going to zero at a critical intensity that depends on the model parameters. In the regime of zero correlation, the particles form a static symmetric circular distribution, following a simple model of aggregation. Interestingly, for intermediate values of the force intensity, different patterns emerge, with the particles spiraling or splitting in two clusters centered at opposite sides of the stimulus’ location. The spiral and two-cluster patterns are stable and active. The two clusters slowly rotate around the source while exchanging particles, or separate and collide repeatedly, depending on the parameters.

show abstract

“…Distributed control and cooperation of multi-agent systems has been a challenging problem within the fields of Robotics and Artificial Intelligence, encouraging exploration of neuroscience-inspired paradigms. Based on recent discoveries in neuroscience [25,26], [24] developed a novel theory for the control of self-organized multi-agent systems within a two-dimensional environment simulating responses to homogeneous data streams based on visual cues. This framework, coined NeuroSwarms, was intended for autonomous swarm control driven by synaptic learning rules that treat multiagent groups analogically to neural network models of spatial cognitive circuits in rodents.…”

Section: Theoretical Background Of Neuroswarmsmentioning

confidence: 99%

Bayesian optimization of distributed neurodynamical controller models for spatial navigation

Hadžić¹,

Hwang²,

Zhang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Dynamical systems models for controlling multi-agent swarms have demonstrated potential advances toward resilient and decentralized spatial navigation algorithms. The bottom-up mechanisms of spatial self-organization in such models can produce complex or chaotic behaviors. For example, we previously introduced the NeuroSwarms controller, in which agent-based interactions were modeled by analogy to neuronal network interactions, including spatial attractor dynamics and temporal phase-synchronization, that have been theorized to operate within the hippocampal place-cell circuits of navigating rodents. While this complexity may drive emergent navigational capabilities, it also precludes linear analyses of stability, controllability, and performance that are used to study conventional swarming models. Further, tuning dynamical controllers by hand or grid search is often inadequate due to the potential complexity of objectives, dimensionality of model dynamics and parameters, and computational costs of continuous-time simulation-based sampling. Here, we present a generalizable framework for tuning dynamical controller models of autonomous multi-agent systems based on Bayesian Optimization (BayesOpt). Our approach utilizes a task-dependent objective

show abstract

Cognitive swarming: an approach from the theoretical neuroscience of hippocampal function

Cited by 10 publications

References 44 publications

Cognitive swarming in complex environments with attractor dynamics and oscillatory computing

Cognitive swarming in complex environments with attractor dynamics and oscillatory computing

Synchronization and spatial patterns in forced swarmalators

Bayesian optimization of distributed neurodynamical controller models for spatial navigation

Contact Info

Product

Resources

About