Figure 3.

LIP neurons reflect evidence from the attended patch of motion. A, Activity of 109 neurons studied in the cued attention task when the upper patch (top) and lower patch (bottom) were cued as informative (combined data from Monkeys Dm and Np). Responses are detrended by neuron, via subtraction of the mean firing rate, as function of time, on the lowest coherences (0% and ±3.2%). Errors on non-zero coherences are excluded. The neurons reflect the formation of decision from information derived from the upper and lower visual field. B, Activity of 36 neurons studied in the variable location task when the motion patch appeared in the upper (top) or lower (bottom) location. Combined activity of neurons from Monkeys Ap and Dz only, as neurons from Monkey Dm recorded in this task did not show dependence on the strength of motion. We report statistics from this task both with and without neurons from Monkey Dm. C, D, Comparison of responses in the cued attention task when motion patches had the same (C) or opposite (D) directions. The graphs use the combined data from both monkeys without detrending.