Figure 6.
Signaling the stronger stimulus across the OTi-d space map. A, Experimental protocol. Pairs of Sin and Sout stimuli of unequal strengths are presented in two stimulus conditions: win, with the stronger stimulus inside the receptive field; and lose, with the weaker one inside. Such tests were performed at different values of relative stimulus strength achieved by fixing the strength of the stronger stimulus and systematically varying the strength of the weaker one; Sin and Sout were looming visual dots. For instance, when the strength of the stronger stimulus was chosen to be 9°/s, the win conditions corresponding to the different relative strengths of 2, 4, 6, and 8°/s were (9, 7°/s), (9, 5°/s), (9, 3°/s), and (9, 1°/s), respectively; pairs indicate Sin and Sout strengths. The fixed strength of the stronger stimulus was chosen such that it evoked at least 50% of the maximal response for that unit (same as the criterion used for choosing Sin strength in Fig. 1). B, Pooled responses (see Materials and Methods) in the win (filled purple circles) and lose (open purple circles) conditions from 20 units with gradual CRPs (left) and 13 units with switch-like CRPs (right). The average strength of the stronger stimulus was 8°/s for both types of CRPs. C, Experimental protocol for measurements of responses to Sin alone. Win and lose conditions were tested in the absence of the Sout stimulus randomly interleaved with the tests in A. Each row represents the stimulus condition that tests the condition in the corresponding row in A but in the absence of Sout. D, Pooled responses in the win (filled gray circles) and lose (open gray circles) conditions from the same units with gradual (left, n = 20) and switch-like (right, n = 13) CRPs, as in A; obtained using the same procedure. E, Discriminability (d′; see Materials and Methods) of the stronger stimulus as a function of the relative stimulus strength for units with gradual (left) and switch-like (right) CRPs. Purple, Data from protocol in A; gray, data from protocol in C. Thick line in right denotes the binary discrimination signal. SEM was estimated by a standard jackknife procedure (Efron and Tibshirani, 1994) and was, in some cases, smaller than the size of the symbol used to indicate the mean. F, MI (see Materials and Methods) quantifying the change in discriminability (d′) of the stronger stimulus attributable to inhibitory competitive interactions. Left, Units with gradual CRPs; right, units with switch-like CRPs. Dashed line shows MI = 0. Data show mean ± SEM; SEM estimated by a standard bootstrap procedure (Efron and Tibshirani, 1994) with 1000 resamplings. In some cases, the SEM is smaller than the size of the symbol used to indicate the mean. * indicates significance at the 0.05 level (t tests with Holm–Bonferroni correction).