E/I imbalance and internal noise cause weak neural representations and face recognition challenges in ASD

Individuals with Autism Spectrum Disorder (ASD) are known for their socio-communicative challenges, including face recognition. It is unclear, however, about its neurocomputational basis and links to neurobiological factors, such as an imbalance of excitatory and inhibitory signals (E/I imbalance) or excessive internal noise (IN). This study employed Convolutional Neural Network (CNN) models to simulate face recognition in typical populations and ASD based on the claims of E/I imbalance and IN theories. We demonstrated that CNN models with non-optimal ReLU slopes or noisy activations yielded poorer performance in face recognition and exhibited atypical neural representations of faces. Overall, simulations based on the E/I imbalance theory seem to encompass a broader range of behavioral and neural profiles in ASD. Our theory-driven approach used CNN models to test neurobiological theories, and our results provide causal evidence on a potential mechanism by which neurobiological factors influence face recognition in ASD.