Cebra

Cebra is an AI technology that compresses time-series data to reveal patterns and differences. It is especially useful when exploring neural data from mice and rats, such as visual cortex or hippocampus activity. With Cebra, it is possible to track movements and uncover complex dynamics, all while using both behavioural and neural data in a self-supervised and learnable way. This leads to highly reliable results in a variety of tasks and behaviours, even across different species. Furthermore, Cebra makes it possible to analyze both single-session and multiple-session datasets without any need for labeling. The software and accompanying paper can be found on arxiv and Github respectively.