Welcome to PsyNeuLink
PsyNeuLink is an open-source, software environment written in Python, and designed for the needs of
neuroscientists, psychologists, computational psychiatrists and others interested in learning about and building
models of the relationship between brain function, mental processes and behavior.
PsyNeuLink can be used as a “block modeling environment”, in which to construct, simulate, document, and exchange
computational models of neural mechanisms and/or psychological processes at the subsystem and system levels.
A block modeling environment allows components to be constructed that implement various, possibly disparate
functions, and then link them together into a system to examine how they interact. In PsyNeuLink, components are
used to implement the function of brain subsystems and/or psychological processes, the interaction of which can then
be simulated at the system level.
The purpose of PsyNeuLink is to make it as easy as possible to create new and/or import existing models, and
integrate them to simluate system-level interactions. It provides a suite of core components for
implementing models of various forms of processing, learning, and control, and its Library includes examples that
combine these components to implement published models. As an open source project, its suite of components is meant
to be enhanced and extended, and its library is meant to provide an expanding repository of models, written in a
concise, executable, and easy to interpret form, that can be shared, and compared by the scientific
- open source, freeing users of the costs or restrictions associated with proprietary software.
- computationally general – it can be used to implement, seamlessly integrate, and simulate interactions among
disparate components that vary in their granularity of representation and function (from individual neurons or
neural populations to functional subsystems and abstract cognitive functions) and at any time scale of execution.
- integrative – it provides a standard and accessible environment for model comparison, sharing, and documentation;
- extensible – it has an interface (API) that allows it to be used with other powerful tools for implementing
individual components, such as:
* Neuron (biophysically realistic models of neuronal function)
* TensorFlow (ODE’s, deep learning);
* Emergent (broad class of neurally-plausible connectionist models);
* ACT-R (symbolic, production system models).
PsyNeuLink is alpha software, that is still being actively developed. Although it is useable, and most of the
documented functionality is available, some features may not yet be fully implemented and/or subject to
modification. Please report any bugs and/or suggestions for development to email@example.com.
What PsyNeuLink is NOT
PsyNeuLink is well suited to the creation of simple to moderately complex models, and for the integration of
disparate existing models into a single, integrated system in which interactions among them can be examined.
While it is fully general, and can be used to implement virtually any type of model, in its current form it is
less well suited to certain kinds of efforts, that involve massively large computations and/or specialized functions
and data types that it currently does not support, such as:
- extensive model fitting
- large scale simulations
- biophysically-realistic models of individual neurons
Other packages that are better suited to such applications are:
TensorFlow (for neural network models);
HDDM (for Drift Diffusion Models);
ACT-R (for production system models);
and Nengo (for biophysically-realistic models of neuronal function).
These packages are good for elaborate and detailed models of a particular form.
In contrast, the focus in designing and implementing PsyNeuLink has been to make it as flexible and easy to use as
possible, with the ability to integrate components constructed in other packages (including some of the ones listed
above) into a single functioning system. These are characteristics that are often (at least in the initial
stages of development) in tension with efficiency (think: interpreted vs. compiled). Three priorities for continued
development are the acceleration of PsyNeuLink, using just-in-time compilation methods, parallelization and adaptation
to FPGA hardware; integration of tools for model analysis (such as those implemented in HDDM); and the implementation
of a graphic interface for the construction of models and realtime display of their execution.
PsyNeuLink is written in Python, and conforms to the syntax, coding standards and modular organization shared by
most Python packages. Basics and Sampler provides an orientation to PsyNeuLinks Components, some examples of what
PsyNeuLink models look like, and some of its capabilities. Quick Reference provides an overview of how PsyNeuLink is
organized and some of its basic principles of operation. The Tutorial provides an interactive guide to the
construction of models using PsyNeuLink. Core contains the fundamental objects used to build PsyNeuLink models, and
Library contains extensions, including speciality components, implemented compositions, and published models.
PsyNeuLink is compatible with python versions >= 3.5, and is available through PyPI:
All prerequisite packages will be automatically added to your environment.
If you downloaded the source code, navigate to the cloned directory in a terminal,
switch to your preferred python3 environment, then run
Lists of required packages for PsyNeuLink, developing PsyNeuLink, and running the PsyNeuLink tutorial are also
stored in pip-style
tutorial_requirements.txt in the source code.
If you have trouble installing the package, or run into other problems, please contact firstname.lastname@example.org.
Download PsyNeuLink Tutorial.ipynb
PsyNeuLink includes a
tutorial, that provides examples of how to create basic Components
in PsyNeuLink, and combine them into Processes and a System. The examples include construction of a simple
decision making process using a Drift Diffusion Model, a neural network model of the Stroop effect, and a
backpropagation network for learning the XOR problem.
The tutorial can be run in a browser by clicking the badge below, or this link.
To run the tutorial locally, you must run python 3.5 and install additional packages:
pip install psyneulink[tutorial]
or if you downloaded the source:
To access the tutorial, make sure you fulfill the requirements
mentioned above, download the
tutorial notebook, then run the terminal command
Once the notebook opens in your browser, navigate to the location where you saved the tutorial notebook, and
click on “PsyNeuLink Tutorial.ipynb”.
- Allie Burton, Princeton Neuroscience Institute, Princeton University
- Jonathan D. Cohen, Princeton Neuroscience Institute, Princeton University
- Peter Johnson, Princeton Neuroscience Institute, Princeton University
- Justin Junge, Department of Psychology, Princeton University
- Kristen Manning, Princeton Neuroscience Institute, Princeton University
- Kevin Mantel, Princeton Neuroscience Institute, Princeton University
- Markus Spitzer, Princeton Neuroscience Institute, Princeton University
- Jan Vesely, Department of Computer Science, Rutgers University
- Changyan Wang, Princeton Neuroscience Institute, Princeton University
- Nate Wilson, Princeton Neuroscience Institute, Princeton University
With substantial and greatly appreciated assistance from:
- Abhishek Bhattacharjee, Department of Computer Science, Rutgers University
- Mihai Capota, Intel Labs, Intel Corporation
- Bryn Keller, Intel Labs, Intel Corporation
- Garrett McGrath, Princeton Neuroscience Institute, Princeton University
- Sebastian Musslick, Princeton Neuroscience Institute, Princeton University
- Amitai Shenhav, Cognitive, Linguistic, & Psychological Sciences, Brown University
- Michael Shvartsman, Princeton Neuroscience Institute, Princeton University
- Ben Singer, Princeton Neuroscience Institute, Princeton University
- Ted Willke, Intel Labs, Intel Corporation