The Self as an Image: Appearance and Belief in Visual Representations of One’s Own Face

Self-representations are central to cognition, yet their internal structure and visual fidelity remain poorly understood. Here, we combine behavioral and computational approaches to recover and assess visual self-face representations from both perception and memory. Participants provided similarity judgments for self, familiar, and unfamiliar faces, which were compared against the representational structure of artificial neural networks (ANNs). We further applied behavior-based image reconstruction to visualize internal representations. Discriminative ANNs aligned more closely with self-face similarity judgments than generative models, suggesting that self-representations emphasize high-level, identity-invariant features while sacrificing lower-level visual detail. Image reconstruction confirmed that self-face representations can be recovered from both perception and memory, though with reduced fidelity relative to familiar and unfamiliar faces. Exploratory analyses further indicated how self-beliefs may modulate these representations: higher self-concept clarity was associated with more accurate representations and self-attractiveness ratings with more attractive ones. Together, these findings indicate that self-face representations constitute a coherent but systematically distorted internal image, shaped by both objective appearance and self-beliefs. By extending image reconstruction to the domain of self-representations, this work provides a novel methodological framework for quantifying their content and underlying distortions.