Cebra

The main purpose of CEBRA is to map behavioural actions to neural activity in order to gain a better understanding of the underlying neural dynamics during adaptive behaviours. The machine learning tool uses non-linear techniques to create consistent and high-performance latent spaces from joint behavioural and neural data recorded simultaneously. It can generate consistent latent embeddings which can be used for hypothesis testing, discovery-driven analysis, kinematic feature mapping, and rapid high-accuracy decoding of natural movies from visual cortex.

CEBRA has been validated for its accuracy and efficacy on calcium and electrophysiology datasets, as well as across sensory and motor tasks and in simple or complex behaviours across species. It can be used with single or multi-session datasets without any labels required. The code is available on GitHub while the pre-print is available on arxiv.org. Therefore, neuroscientists looking for an effective way of analyzing and decoding behavioural and neural data can make use of this powerful tool to uncover underlying neural representations.