Results showing that the main attention and isolated object decoding accuracies generalize across position. For this analysis we trained the classifier using data from one location and then tested the classifier with data from when either the isolated object was shown at a different location (blue trace), or when the attended or nonattended object was shown at a different location (red and green traces). The results are averaged over training and testing at all locations. The upper plot show the results from the 16 object stimulus set (data combined from both monkeys), and the lower plots shows the results from the 7 object stimulus set (data from monkey 2).
