Abstract
Humans guide visual attention to different targets as they move through different environments (i.e., going from a school zone to a construction zone while driving). To monitor how observers relearn which target features to prioritize across such contextual shifts, we had observers perform a visual search task where target shapes are presented in different colors with distinct probabilities (e.g., 33%, 26%, 19%, 12%, and 5%). Then, we periodically shifted the color-to-probability mapping. Here we show that people rapidly relearn complex attentional priority maps, and that they show no evidence for interference as they shift from old to new attentional sets, even when the new color mappings are in direct conflict with the old. Our findings suggest that learning to control attention may be interference free, unlike other forms of visual learning.