Sarah E. Purnell¹ (), Daniel D. Thayer¹, Thomas C. Sprague¹; ¹University of California, Santa Barbara

Foreknowledge of task-relevant features and locations guides attention during visual search, whereby information within relevant locations is strengthened while extraneous information at irrelevant locations is minimized (Kastner et al., 1999; Theeuwes, 2004). It is believed that this, in part, is due to interactions between top-down ‘relevance maps’ and ‘feature dimension maps’ in feature-selective regions encoding salient locations (Thayer & Sprague, 2023; 2025). Based on this account, we would predict that different visual features (e.g., color, motion, shape) defining task-relevant and salient stimuli, act equivalently, however, spatial attention may differentially affect interference from color- and motion-defined distractors (Leonard et al., 2015; Liu & Mance, 2011). We tested this by having participants covertly search an 8-item array for a unique target item in a prespecified feature dimension (e.g., color, motion, shape) and indicate the orientation of a line (horizontal or vertical) inside the target. Most trials were preceded by a valid, central cue indicating the task-relevant hemifield of the search array (top, left, bottom, or right; 4 of 8 items), where the target always appeared. On half of all trials, a singleton distractor in a task-irrelevant dimension (e.g., color distractor in a motion-defined search) could appear within or outside the cued hemifield. Singleton distractors were operationalized as the opposite color, motion direction, or shape (Li et al., 2019) of all other items in the array along a 360-degree continuous feature space. Cueing spatial relevance reduced distractor interference at task-irrelevant locations, but these effects differed substantially across variations in target- and distractor-defined feature dimensions. These results indicate that while top-down goals may attenuate distractor-induced capture to guide visual search, this attenuation is not uniform across all types of distraction. This motivates future studies on how the neural mechanisms mediating task relevance and salience may sculpt neural representations in feature-selective regions of the cortex.

Acknowledgements: Funding: National Eye Institute R01-EY035300