Abstract:Abstract Many cognitive processes involve transformations of distributed representations in neural populations, creating a need for population-level models. Recurrent neural network models fulfill this need, but there are many open questions about how their connectivity gives rise to dynamics that solve a task. Here, we present a method for finding the connectivity of networks for which the dynamics are specified to solve a task in an interpretable way. We apply our method to a working memory task by synthesizing a network that implements a drift-diffusion process over a ring-shaped manifold. We also use our method to demonstrate how inputs can be used to control network dynamics for cognitive flexibility and explore the relationship between representation geometry and network capacity. Our work fits within the broader context of understanding neural computations as dynamics over relatively low-dimensional manifolds formed by correlated patterns of neurons. Author Summary Neurons in the brain form intricate networks that can produce a vast array of activity patterns. To support goal-directed behavior, the brain must adjust the connections between neurons so that network dynamics can perform desirable computations on behaviorally relevant variables. A fundamental goal in computational neuroscience is to provide an understanding of how network connectivity aligns the dynamics in the brain to the dynamics needed to track those variables. Here, we develop a mathematical framework for creating recurrent neural network models that can address this problem. Specifically, we derive a set of linear equations that constrain the connectivity to afford a direct mapping of task-relevant dynamics onto network activity. We demonstrate the utility of this technique by creating and analyzing a set of network models that can perform a simple working memory task. We then extend the approach to show how additional constraints can furnish networks whose dynamics are controlled flexibly by external inputs. Finally, we exploit the flexibility of this technique to explore the robustness and capacity limitations of recurrent networks. This network synthesis method provides a powerful means for generating and validating hypotheses about how task-relevant computations can emerge from network dynamics.

Flexible multitask computation in recurrent networks utilizes shared dynamical motifs

Synergistic pathways of modulation enable robust task packing within neural dynamics

Emergence and reconfiguration of modular structure for synaptic neural networks during continual familiarity detection

Emergence and reconfiguration of modular structure for artificial neural networks during continual familiarity detection

Engineering recurrent neural networks from task-relevant manifolds and dynamics

Task representations in neural networks trained to perform many cognitive tasks

Structured flexibility in recurrent neural networks via neuromodulation

Neural Computations in A Dynamical System with Multiple Time Scales

Modular Boundaries in Recurrent Neural Networks

State dependent computation using coupled recurrent networks

Analysis of Structure and Dynamics in Three-Neuron Motifs

A Geometric Framework for Understanding Dynamic Information Integration in Context-Dependent Computation

Flexible task abstractions emerge in linear networks with fast and bounded units

Modular representations emerge in neural networks trained to perform context-dependent tasks

Modular Continual Learning in a Unified Visual Environment

Growing Brains: Co-emergence of Anatomical and Functional Modularity in Recurrent Neural Networks

Shaping dynamical neural computations using spatiotemporal constraints

Beyond modularity: Fine-scale mechanisms and rules for brain network reconfiguration

Building compositional tasks with shared neural subspaces

Parallel Learning by Multitasking Neural Networks

Biologically plausible models of cognitive flexibility: merging recurrent neural networks with full-brain dynamics