Background <p>Intelligence, empathy, systemizing, and autistic traits are psychological dimensions that vary continuously across the population. Autism spectrum disorder (ASD) may reflect extreme expressions of these traits, yet their interdependencies and difference between typically developing (TD) children remain poorly understood.</p> Methods <p>We recruited 186 children with ASD and 364 TD children aged 6–12 years. Intelligence was assessed using the Wechsler Intelligence Scale for Children–Fourth Edition, while empathy, systemizing, and autistic traits were measured using validated Chinese versions of the Empathy Quotient, Systemizing Quotient, and Autism Spectrum Quotient. Undirected partial correlation networks and Bayesian networks were estimated separately for each group, adjusting for demographic covariates, including gender, parental education level, and household income.</p> Results <p>Across groups, cognitive empathy consistently emerged as the central hub in both undirected and Bayesian network analyses. In ASD, autistic traits formed a densely interconnected cluster with limited cross-domain links, while in TD children they were more flexibly integrated with systemizing and other domains. Intelligence formed a cohesive yet relatively independent subnetwork, though in ASD verbal comprehension was directly predicted by social communication competence. The estimated networks demonstrated robust accuracy and acceptable stability.</p> Conclusion <p>These findings reveal both shared and distinct patterns of network organization among intelligence, empathy, systemizing, and autistic traits in children with and without ASD. Cognitive empathy may represent a potential target for future intervention research, while systemizing tendencies and cognitive strengths may also inform individualized support strategies in ASD.</p>