Blindfold learning of a neural metric

To understand how neural signals encode sensory stimuli, we often require to know both the true stimulus and the neural response. The brain, however, only has access to the noisy and correlated response of sensory neurons. How can it map out these responses in a way that is pertinent for stimulus interpretation? I will show how to build an accurate distance map of responses solely from the structure of the population activity of retinal ganglion cells, based on a Temporal Restricted Boltzmann Machine trained to describe the spatiotemporal structure of the population activity. This metric outperforms existing neural distances at discriminating pairs of stimuli that are barely distinguishable. The proposed method provides a generic and biologically plausible way to learn to associate similar stimuli based on their spiking responses, without any other knowledge of these stimuli.