Cebra

CEBRA Overview

CEBRA is a machine-learning method developed by Steffen Schneider, Jin Hwa Lee, and Mackenzie Mathis at EPFL & IMPRS-IS. This innovative tool is designed to compress time series in a way that uncovers hidden structures in the variability of data. It is particularly effective with behavioural and neural data recorded simultaneously, and can decode activity from the visual cortex of the mouse brain to reconstruct a viewed video. CEBRA is a powerful tool for neuroscience, offering both consistent and high-performance latent spaces, and can be used for hypothesis testing or label-free.

CEBRA Highlights

• CEBRA can be used to compress time series, revealing hidden structures in the variability of data.
• It is effective with behavioural and neural data recorded simultaneously, and can decode activity from the visual cortex of the mouse brain to reconstruct a viewed video.
• CEBRA offers both consistent and high-performance latent spaces, and can be used for hypothesis testing or label-free.