Using Neural Networks to Uncover the Relationship between Highly Variable Behavior and EEG during a Working Memory Task with Distractors

Value-driven attention capture (VDAC) occurs when previously rewarded stimuli and impair goal-directed behavior. In a working memory (WM) task with VDAC-related distractors, we observe behavioral variability both within across individuals. Individuals differ in their ability to maintain relevant information ignore distractions. These cognitive components shift over time changes motivation attention, making it difficult identify underlying neural mechanisms of individual differences. this study, develop the first participant-specific feedforward network models reaction from data during VDAC WM task. We used short epochs electroencephalography (EEG) 16 participants (NN) RT aimed at understanding VDAC. Using general linear (GLM), identified 20 EEG features predict (r=0.53±0.08). The model was compared NN model, which improved predicted trial-by-trial for all (r=0.87±0.04). found that right frontal gamma-band activity fronto-posterior functional connectivity alpha, beta, gamma bands explain Our study shows can link highly variable behavior potential new targets neuromodulation interventions.