# Scalable Spike Source Localization in Extracellular Recordings using Amortized Variational Inference

## TL;DR

This paper introduces a scalable Bayesian approach using amortized variational inference for localizing neuron sources in extracellular recordings, improving accuracy over traditional heuristic methods.

## Contribution

It presents a novel variational autoencoder-based model for efficient, scalable spike source localization in high-density neural recordings.

## Key findings

- More accurate than heuristic localization methods
- Improves spike sorting performance
- Effective on both simulated and real data

## Abstract

Determining the positions of neurons in an extracellular recording is useful for investigating functional properties of the underlying neural circuitry. In this work, we present a Bayesian modelling approach for localizing the source of individual spikes on high-density, microelectrode arrays. To allow for scalable inference, we implement our model as a variational autoencoder and perform amortized variational inference. We evaluate our method on both biophysically realistic simulated and real extracellular datasets, demonstrating that it is more accurate than and can improve spike sorting performance over heuristic localization methods such as center of mass.