Characterizing the shared signals of face familiarity: long-term acquaintance, voluntary control, and concealed knowledge

Using cross-experiment multivariate classification of EEG patterns, in a recent study we found evidence for a shared familiarity signal for faces, patterns of neural activity that successfully separate trials for familiar and unfamiliar faces across participants and modes of familiarization. Here, our aim was to expand upon this research to further characterize the spatio-temporal properties of this signal. By utilizing the information content present for incidental exposure to personally familiar and unfamiliar faces, we tested how the information content in the neural signal unfolds over time under different task demands - giving truthful or deceptive responses to photographs of genuinely familiar and unfamiliar individuals. For this goal, we re-analyzed data from two previously published experiments using within-experiment leave-one-subject-out and cross-experiment classification of face familiarity. We observed that the general face familiarity signal, consistent with its previously described spatio-temporal properties, is present for long-term personally familiar faces under passive viewing, as well as for acknowledged and concealed familiarity responses. Also, central-posterior regions contain information related to deception. We propose that signals in the 200-400 ms window are modulated by top-down task-related anticipation, while the patterns in the 400-600 ms window are influenced by conscious effort to deceive. To our knowledge, this is the first report describing the representational dynamics of concealed knowledge for faces. HighlightsO_LIPrevious studies found a robust EEG effect for face-familiarity in the 200-600 ms post-stimulus range. C_LIO_LIThis neural pattern was found to be shared across participants and modes of familiarization. C_LIO_LIWe used incidental exposure as a template to probe information content for acknowledged and concealed familiarity C_LIO_LIThe shared familiarity signal is modulated differentially in early (200-400 ms) and late (post-400 ms) windows C_LIO_LICross-experiment classification is a promising tool to investigate how cognitive processes unfold under different conditions C_LI