Explaining how neuronal activity gives rise to behaviour and cognition is a central goal of cognitive neuroscience. With the proliferation of larger neuronal datasets, there have been various attempts to abstract representations of the neuronal data. Some methods consider behavioural decoding to be important while other unsupervised methods like PCA and autoencoder disregard behaviour altogether. Here, we propose an architecture to learn cognitive state representations which preserve information of both the dynamics and behaviour. We present a neural network implementation (BunDLe Net) and apply it on calcium imaging neuronal data of the roundworm C. elegans. Our method reveals clear orbit-like trajectories which are recurrent and structured. It also outperforms conventional methods in the field such as PCA, autoencoders and autoregressors with regards to the dynamical predictability and behavioural decoding accuracy.

Top