Spiking activity in the ventral visual pathway under delayed match-to-sample task

The hierarchical view of the ventral object recognition pathway is primarily based on feedforward mechanisms, starting from a fixed basis set of object primitives and ending on a representation of whole objects in the inferotemporal cortex. Here we provide a different view. Rather than being a fixed “labeled line” for a specific feature, neurons are continually changing their stimulus selectivities on a moment-to-moment basis, as dictated by top-down influences of object expectation and perceptual task. Here we also derive the selectivity for stimulus features from an ethologically curated stimulus set, based on a delayed match-to-sample task, that finds components that are informative for object recognition., We implanted microelectrode arrays (Blackrock) in different parts of the ventral visual pathway in the left hemisphere. In animal MJ we implanted four arrays, three 6 x 6 and one 4 x 5, for a total of 128 electrodes with 400 µm spacing between electrodes. All electrodes were 1 mm long. One 6 x 6 array was implanted in opercular V1 representing 2° eccentricity in a lower visual field quadrant; the others were implanted at three positions along the prelunate gyrus. Animal MR was implanted with three arrays: one 6 x 8 array in opercular V1 representing 2° eccentricity in the lower visual field quadrant, and two 5 x 8 arrays in the prelunate gyrus. With the arrays we were able to record single unit and multiunit action potentials and local field potentials. Multiunit activity was manually sorted offline with SpikeSort software (Plexon). Sorted spikes are binned in 1ms bins.
The animals were given an initial cue or sample consisting of a full object for 500 msec, a mask consisting of a jumbl..., , [https://doi.org/10.5061/dryad.fttdz091d](https://doi.org/10.5061/dryad.fttdz091d)

## Description of the data and file structure

Each file represents one dataset. The .mat file is openable with Matlab. The file name represents the acquisition date (YYMMDD) followed by the anima's identification (either J or R). Each row on the dataset represents a trial. Columns represent the electrode number, unit (electrodes can record more than one cell, which we sorted manually), cue stimulus, cue presentation time (msec) , target stimulus, target presentation time (msec), behavior (1 for correct answer, 0 for incorrect answer and -1 for time out), and the time series of spike count (1ms bins). The images named in the file are available on Zenodo, which has the cue images, scrambled imaged and target images. The sequence of the trial is cue followed by scrambled followed by target.

The trials all have different lengths, so it is important that all trials be aligned according to the portion you wi...,