The same conclusion was also supported by the distribution of saccades to the different types of stimuli. In the search array with 20 stimuli, the average percentages of total stimuli comprised by the target, by distracters that shared the target color (share-color), by distracters that shared the target shape (share-shape), and by distracters that shared no target features (no-share) were 5% (1 of 20), 10% (2 of 20), 10% (2 of 20), and 75% (15 of 20), respectively. If monkeys made saccades

to stimuli without using the target features to guide their search, the percentage of saccades to each type of stimulus should match the stimulus frequency. Instead, the percentage of saccades to these four types of stimuli were 34.3%, 14.1%, 12.3%, and 39.3%, respectively, for monkey G, and 34.7%, 20.1%, 8.7%, and 36.4%, respectively, KRX-0401 supplier for monkey L. Thus, the animals made eye movements to the targets and distracters that shared target features more often than to no-share distracters expected by their frequency in the array, supporting the Tyrosine Kinase Inhibitor Library research buy idea that the monkeys used the target features to guide their search. We recorded 134 sites with visual responses

in the FEF and 136 sites with visual responses in V4 in the two monkeys (Figure 1C). The results were qualitatively similar in both monkeys and were therefore

combined. RFs were mapped in a memory-guided saccade task (see Experimental Procedures). On average, the RFs of FEF sites covered 4.5 ± 0.16 stimuli in the search array. Figure S1E shows responses of a FEF site during this task. To isolate the feature-based attention effect, we sorted fixations during the search period according to the category of stimuli in the RF: “target,” “share-color,” “share-shape,” and “no-share” distracter (Figure 1B). In the target fixations, the target was in the RF. In the share-color and share-shape fixations, a distracter was in the RF, and it shared the target see more color or shape, respectively, and in the no-share fixations the distracter in the RF shared no target features. To isolate the effects of feature attention from those of spatial attention, we only included fixations in which the following saccade was made away from the RF for this analysis, e.g., a share-color fixation was one where a share-color distracter was in the RF, but the saccade was made to a stimulus outside of the RF. We also matched the stimuli in the RF across comparison conditions, so there was no difference in the stimuli themselves across attention conditions (see Experimental Procedures).