A gaze is a complex stimulus that provides valuable social information during human interactions. It shares the ability to orient attention with other nonsocial stimuli, such as arrows, but, still, gaze generates unique effects. A clear example was found using a spatial interference task (Marotta et al, 2018). Participants had to identify the direction—left or right—indicated by arrows or eyes located either on the right or left side. Arrows, like other nonsocial stimuli, elicited a standard congruency effect, with faster responses to congruent location-direction trials than to incongruent ones. In contrast, gaze produced a reversed effect with faster responses to incongruent than to congruent trials. Socio-cognitive components of gaze processing could underlie this dissociation. Similar to a joint attention episode in everyday life, gaze would direct attention to a potentially relevant item. In congruent trials (i.e., left-located eyes looking to the left), gaze would orient attention outward, causing a slower response compared to incongruent trials where gaze would look to the central fixation point. With this study, we aimed at better understanding the underlying mechanisms of the reversed congruency effect of gaze by modifying the task—adding a bicolor frame—and its instructions. One group of 25 adults performed the task by identifying stimuli direction, as in previous studies. Another group had to identify the color to which stimuli were directed. We expected to find analogous congruency effects for arrows and gaze in the latter group since both stimuli would direct attention to color and held at this location, preventing gaze from seeking a potential attentional target during congruent trials. Although we found a reduced gaze effect, overall results were similar in both groups so the manipulation of instructions did not have the anticipated effect. The limitations of the study and the need for further investigations were discussed.