Abstract
Task-trained artificial recurrent neural networks (RNNs) provide a computational modeling framework of increasing interest and application in computational, systems, and cognitive neuroscience. RNNs can be trained, using deep-learning methods, to perform cognitive tasks used in animal and human experiments and can be studied to investigate potential neural representations and circuit mechanisms underlying cognitive computations and behavior. Widespread application of these approaches within neuroscience has been limited by technical barriers in use of deep-learning software packages to train network models. Here, we introduce PsychRNN, an accessible, flexible, and extensible Python package for training RNNs on cognitive tasks. Our package is designed for accessibility, for researchers to define tasks and train RNN models using only Python and NumPy, without requiring knowledge of deep-learning software. The training backend is based on TensorFlow and is readily extensible for researchers with TensorFlow knowledge to develop projects with additional customization. PsychRNN implements a number of specialized features to support applications in systems and cognitive neuroscience. Users can impose neurobiologically relevant constraints on synaptic connectivity patterns. Furthermore, specification of cognitive tasks has a modular structure, which facilitates parametric variation of task demands to examine their impact on model solutions. PsychRNN also enables task shaping during training, or curriculum learning, in which tasks are adjusted in closed-loop based on performance. Shaping is ubiquitous in training of animals in cognitive tasks, and PsychRNN allows investigation of how shaping trajectories impact learning and model solutions. Overall, the PsychRNN framework facilitates application of trained RNNs in neuroscience research.
Highlights
Studying artificial neural networks (ANNs) as models of brain function is an approach of increasing interest in computational, systems, and cognitive neuroscience (Kriegeskorte, 2015; Yamins and DiCarlo, 2016; Richards et al, 2019)
Package structure To serve our objectives of accessibility, extensibility, and reproducibility, we divided the PsychRNN package into two main components: the Task object and the Backend (Fig. 1)
We demonstrate how PsychRNN can specify an recurrent neural network (RNN) model, train it to perform a task of neuroscientific interest, here, a two-alternative forced-choice perceptual discrimination task (Roitman and Shadlen, 2002), and return behavioral readout from output units and internal activity patterns of recurrent units (Fig. 2)
Summary
Studying artificial neural networks (ANNs) as models of brain function is an approach of increasing interest in computational, systems, and cognitive neuroscience (Kriegeskorte, 2015; Yamins and DiCarlo, 2016; Richards et al, 2019). ANNs comprise many simple units, called neurons, whose synaptic connectivity patterns are iteratively updated via deep-learning methods to optimize an Received September 28, 2020; accepted December 2, 2020; First published December 16, 2020. The authors declare no competing financial interests. D.B.E., J.T.S., D.B., and A.A. developed software. D.B.E., J.T.S. and J.D.M. wrote the paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
