Activation profiles across neural feature dimension maps guide attention during visual search

Image-salient distractors and goal-relevant targets simultaneously compete for our attention. To resolve competition among items in the visual field, maps that preferentially respond to specific feature dimensions (e.g., color or motion) index salient and relevant locations based on their preferred feature, which is subsequently integrated into a feature-agnostic priority map, where the most important location guides attention. Even though feature dimension maps play a critical role in guiding attention, it is unclear how activation profiles in these neural maps resolve competition between task-relevant and irrelevant, but salient, items. Here, we used a visual search task to evaluate how relevant and salient items compete in neural feature dimension maps. On each trial, participants were cued to search for a target defined by a specific color or motion direction in a subsequent search array containing 8 colorful moving dot stimuli. All items in the array had homogenous features, except for the target item, which differed solely on the cued feature dimension. Occasionally, one of the non-target items was a salient distractor presented with either a different color or motion direction from the other items. We used an inverted encoding model to reconstruct spatial maps from activation patterns in feature-selective retinotopic regions (motion [TO1/TO2] and color [hV4/VO1/VO2] maps). Both targets and distractors were represented via spatially localized heightened activation in reconstructed spatial maps. Furthermore, we compared trials where the salient distractor likely captured attention (slow search RTs) to those where it was likely ignored (fast RTs). Activation within neural motion dimension maps reflected greater activation of the relevant target relative to the salient distractor when search was fast, and the inverse result when search was slow. These results indicate that neural feature dimension maps are crucial for computing attentional priority and that activation profiles across these maps guide behavior.