Abstract
The probability of selecting a visual stimulus as the goal of an eye movement during efficient visual search depends on the value of that stimulus in the context of the task. We have previously shown that activity in the lateral intraparietal area (LIP) represents a priority map, the activity of which can explain efficient search. It has been suggested that activity in LIP is correlated with the likelihood of reward, set size or global computations of relative reward value during visual decision making task. However, it is not clear how values for saccade goal selection are prioritized while foraging through a cluttered visual scene. We trained two animals to perform a visual foraging task in which they were free to search through potential targets and distractors to identify the target that was loaded with reward. To get the reward they had to fixate the loaded target for 500 ms. The number of targets and distractors varied in each trial. In half of the trials the total number of objects was 10 but the number of potential targets varied among 3, 5 or 7. In the other half of the trials the number of potential targets was 3 but the number of objects varied among 3, 5 or 7. Stimuli were spaced such that when looking at one stimulus, another was in the LIP neuron's receptive field. We find that the probability of making a saccade to a target or a distractor can be precisely predicted as a function of the number of targets and distractors. The predicted function can be used to calculate the trial-by-trial attentional priority, which is highly correlated to the weighted normalized activity of LIP neurons. We suggest that the activity of LIP neurons represents the attentional priority of the goal in context of the task.
NIH Grant EY019273, The McKnight Foundation.