Learning interpretable control inputs and dynamics underlying animal locomotion

TS Mullen, M Schimel, G Hennequin… - The Twelfth …, 2024 - openreview.net
A central objective in neuroscience is to understand how the brain orchestrates movement.
Recent advances in automated tracking technologies have made it possible to document …

Structure-preserving recurrent neural networks for a class of Birkhoffian systems

S Xiao, M Chen, R Zhang, Y Tang - Journal of Systems Science and …, 2024 - Springer
In this paper, the authors propose a neural network architecture designed specifically for a
class of Birkhoffian systems—The Newtonian system. The proposed model utilizes recurrent …

Reducing power requirements for high-accuracy decoding in iBCIs

BM Karpowicz, B Bhaduri… - Journal of Neural …, 2024 - iopscience.iop.org
Objective: Current intracortical brain-computer interfaces (iBCIs) rely predominantly on
threshold crossings (“spikes”) for decoding neural activity into a control signal for an external …

When predict can also explain: few-shot prediction to select better neural latents

K Dabholkar, O Barak - arXiv preprint arXiv:2405.14425, 2024 - arxiv.org
Latent variable models serve as powerful tools to infer underlying dynamics from observed
neural activity. However, due to the absence of ground truth data, prediction benchmarks are …

Learning low-dimensional dynamics from whole-brain data improves task capture

E Geenjaar, D Kim, R Ohib, M Duda, A Kashyap… - arXiv preprint arXiv …, 2023 - arxiv.org
The neural dynamics underlying brain activity are critical to understanding cognitive
processes and mental disorders. However, current voxel-based whole-brain dimensionality …

Universal Differential Equations as a Common Modeling Language for Neuroscience

A ElGazzar, M van Gerven - arXiv preprint arXiv:2403.14510, 2024 - arxiv.org
The unprecedented availability of large-scale datasets in neuroscience has spurred the
exploration of artificial deep neural networks (DNNs) both as empirical tools and as models …

Diffusion-Based Generation of Neural Activity from Disentangled Latent Codes

JD McCart, AR Sedler, C Versteeg, D Mifsud… - arXiv preprint arXiv …, 2024 - arxiv.org
Recent advances in recording technology have allowed neuroscientists to monitor activity
from thousands of neurons simultaneously. Latent variable models are increasingly valuable …

Models of neural circuits as optimally driven dynamical systems

M Schimel - 2024 - repository.cam.ac.uk
Animal brains are composed of large numbers of neurons, whose time-varying activity is
shaped by their recurrent connections.% neurons are connected. Recent advances in …

Deciphering the Geometry of Primary Motor Cortical Manifolds: Observations From Naturalistic Movements and Implications for Intracortical Brain-Computer Interfaces

E Altan - 2023 - search.proquest.com
Each neuron in the primary motor cortex (M1) is like a musician in an orchestra, contributing
to a larger harmony under the constraint of a “neural manifold”—a geometric score …

Uncovering the latent dynamics of whole-brain fMRI tasks with a sequential variational autoencoder

E Geenjaar, D Kim, R Ohib, M Duda, A Kashyap… - … Generative Models for … - openreview.net
The neural dynamics underlying brain activity are critical to understanding cognitive
processes and mental disorders. However, current voxel-based whole-brain dimensionality …