Reproduce “Mesoscopic modeling of hidden spiking neurons” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shuqi Wang, Valentin Schmutz, Guillaume Bellec, Wulfram Gerstner
A neuronally-grounded latent variable model (neuLVM) from EPFL researchers infers spiking neural network parameters and hidden population activity from partial observations, accurately reproducing metastable dynamics and predicting responses to perturbations. This approach makes modeling large, partially observed neural circuits computationally tractable while maintaining biological interpretability.
Key insights
- The aggregate activity of unobserved neurons can be effectively captured by a coarse-grained, mesoscopic description, transforming an intractable problem into a tractable one.
- Deriving latent variables and model parameters directly from SNN principles ensures transparent biological interpretation, allowing for testing mechanistic hypotheses.
- Neuronally-grounded inductive biases enable robust inference of complex dynamics and parameter recovery from recordings with a limited fraction of observed neurons.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.