Computation-through-Dynamics Toolkit: Simulated datasets and quality metrics for dynamical models of neural activity
12 affiliations
- 1Wallace H. Coulter Department of Biomedical Engineering, Emory University and Georgia Institute of Technology, Atlanta, Georgia, United States of America
- 2Center for Machine Learning, Georgia Institute of Technology, Atlanta, Georgia, United States of America
- 3Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States of America
- 4Wu Tsai Neurosciences Institute, Stanford University, Stanford, California, United States of America
- 5Department of Statistics, Stanford University, Stanford, California, United States of America
- 6Allen Institute for Neural Dynamics, Seattle, Washington, United States of America
- 7Department of Neurobiology & Biophysics, University of Washington, Seattle, Washington, United States of America
- 8Département de Neurosciences, Faculté de Médecine, Université de Montréal, Montréal, Canada
- 9MILA, Quebec Artificial Intelligence Institute, Montréal, Quebec, Canada
- 10School of Kinesiology and Health Science, Faculty of Health, York University, Toronto, Ontario, Canada
- 11Department of Electrical Engineering, Stanford University, Stanford, California, United States of America
- 12Department of Neurosurgery, Emory University, Atlanta, Georgia, United States of America
Abstract
A primary goal of systems neuroscience is to discover how ensembles of neurons transform inputs into goal-directed behavior, a process known as neural computation. A powerful framework for understanding neural computation uses neural dynamics – the rules that govern how neural activity evolves over time – to explain how goal-directed input-output transformations occur. As dynamical rules are not directly observable, we need computational models that can infer neural dynamics from recorded neural activity. We typically validate such models using synthetic datasets with known ground-truth dynamics, but unfortunately existing synthetic datasets don’t reflect fundamental features of neural computation and may therefore be poor proxies for neural systems. Further, the field lacks validated metrics for quantifying the accuracy of the dynamics inferred by models. The Computation-through-Dynamics Toolkit (CtDToolkit) addresses these critical gaps by providing: 1) synthetic datasets that reflect computational properties of biological neural circuits, 2) interpretable metrics for quantifying model performance, and 3) a standardized pipeline for training and evaluating models with or without known external inputs. In this manuscript, we demonstrate how CtDToolkit can help guide the development, tuning, and troubleshooting of neural dynamics models. In summary, CtDToolkit provides a necessary framework for model developers to better understand and characterize neural computation through the lens of dynamics. Author summary: Understanding how the brain works requires interpretable accounts of how populations of neurons process information to produce behavior. One powerful approach is to study “neural dynamics,” the patterns of how neural activity evolves over time. Scientists develop computational models to infer these dynamics from neural recordings, but it has been challenging to know when the inferred dynamics are trustworthy. Existing datasets often lack key features of biological neural circuits, and current performance metrics can provide an incomplete picture of model quality.
Related
- VariantFlow: a selective-execution engine for efficient population genomic computation on large variant datasets
- Impact of Simulated Gastrointestinal Digestion on Antiglycoxidant Activity of Lemon Verbena (Aloysia triphylla) Herbal Tea and Characterization of Key Polyphenols via DPPH/MGO Pre-Column HPLC
- Learning Environmental Dynamics: Building Internal Models of a Simulated External Force
- Identification and validation of biomarkers associated with mitochondrial dynamics dysfunction in Parkinson’s disease based on GEO datasets
- Breadboard Neuroscience: Exploring Neuronal Dynamics and Computation with Simple Electronic Circuits
BibTeX
@article{versteeg2026computationthroughdynamics,
title = {{Computation-through-Dynamics Toolkit: Simulated datasets and quality metrics for dynamical models of neural activity}},
author = {Christopher Versteeg and Jonathan D. McCart and Mitchell Ostrow and David M. Zoltowski and Clayton B. Washington and Laura Driscoll and Olivier Codol and Jonathan A. Michaels and Scott W. Linderman and David Sussillo and Chethan Pandarinath},
journal = {PLOS Computational Biology},
year = {2026},
month = oct,
doi = {10.1371/journal.pcbi.1014755},
url = {https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014755},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Computation-through-Dynamics Toolkit: Simulated datasets and quality metrics for dynamical models of neural activity
AU - Christopher Versteeg
AU - Jonathan D. McCart
AU - Mitchell Ostrow
AU - David M. Zoltowski
AU - Clayton B. Washington
AU - Laura Driscoll
AU - Olivier Codol
AU - Jonathan A. Michaels
AU - Scott W. Linderman
AU - David Sussillo
AU - Chethan Pandarinath
PY - 2026
DA - 2026/10/15
JO - PLOS Computational Biology
DO - 10.1371/journal.pcbi.1014755
UR - https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014755
AB - A primary goal of systems neuroscience is to discover how ensembles of neurons transform inputs into goal-directed behavior, a process known as neural computation. A powerful framework for understanding neural computation uses neural dynamics – the rules that govern how neural activity evolves over time – to explain how goal-directed input-output transformations occur. As dynamical rules are not directly observable, we need computational models that can infer neural dynamics from recorded neural activity. We typically validate such models using synthetic datasets with known ground-truth dynamics, but unfortunately existing synthetic datasets don’t reflect fundamental features of neural computation and may therefore be poor proxies for neural systems. Further, the field lacks validated metrics for quantifying the accuracy of the dynamics inferred by models. The Computation-through-Dynamics Toolkit (CtDToolkit) addresses these critical gaps by providing: 1) synthetic datasets that reflect computational properties of biological neural circuits, 2) interpretable metrics for quantifying model performance, and 3) a standardized pipeline for training and evaluating models with or without known external inputs. In this manuscript, we demonstrate how CtDToolkit can help guide the development, tuning, and troubleshooting of neural dynamics models. In summary, CtDToolkit provides a necessary framework for model developers to better understand and characterize neural computation through the lens of dynamics. Author summary: Understanding how the brain works requires interpretable accounts of how populations of neurons process information to produce behavior. One powerful approach is to study “neural dynamics,” the patterns of how neural activity evolves over time. Scientists develop computational models to infer these dynamics from neural recordings, but it has been challenging to know when the inferred dynamics are trustworthy. Existing datasets often lack key features of biological neural circuits, and current performance metrics can provide an incomplete picture of model quality.
SN - 1553-7358
SN - 1553-734X
ER -