According to action control theories, responding to a stimulus leads to binding of stimulus features and response. Repeating any component retrieves the previous information, leading to benefits for full repetitions but interference for partial repetitions. These partial-repetition costs are typically absent in localization tasks, in which participants signal the location of a stimulus. Yet, it has been shown that binding effects can occur in localization performance, if the location feature is processed in a post-selective step before the response is executed. Additionally, it has been shown, that binding effects are reduced in overlearned tasks. Both effects might be mediated by involvement of prefrontal brain structures in task processing. Thus, previous studies have shown the correlation between anterior dorsolateral prefrontal cortex activity and response-response binding effects. In the current study, participants localized targets systematically repeating or changing their location and/or color. Crucially, participants had to click on a spatially congruent key (direct response), or a key diagonally opposite to the target (translated response). During these tasks, neural activity was measured with functional near-infrared spectroscopy. On a behavioral level, we replicated that binding effects were present in the translated condition but absent in the direct condition. Further, neuronal activation was higher in the translated condition than in the direct condition. Also, higher activity, specifically in the anterior region, correlated with greater binding effects. Our results suggest that localization tasks with an arbitrary response mapping need more cognitive processing capacity to execute an action and for this reason, binding effects materialize.