Figure 3.

Category classification across locations. A: We asked how much information about preferred and nonpreferred categories is present in category‐selective regions across locations. Thus, we trained a classifier to distinguish between activation patterns of exemplars from different categories presented in one location, and tested the classifier on activation patterns of the same exemplars in the other location. B: Results of category classification across locations in category‐selective cortex. All regions contained more information about preferred than nonpreferred categories. Interestingly, nearly all regions also contained information about nonpreferred categories. Results represent the mean decoding accuracy ± SEM. The red dashed line indicates classification chance level (50%). C: Results of category classification across locations in low‐level visual cortex. There was no evidence for category information across locations in low‐level visual areas. Results represent the mean decoding accuracy ± SEM. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]