EEG-Based Computational Modeling of Creative Cognition: A Hybrid Framework for Neural Signal Interpretation

Liyan Nii^*

• Yulin University, Yulin, China

The quantitative analysis of EEG signals offers a powerful means to understand the neural basis of creative behavior, which is crucial for advancing fields such as neuroscience, psychology, and education. Creative behavior involves complex cognitive processes, including divergent thinking, ideation, and associative processing, which are reflected in specific patterns of neural activity observable in EEG signals. By identifying and modeling these patterns, researchers can better interpret how creative thinking unfolds over time and design systems that enhance creative potential. While traditional EEG analysis methods such as Granger causality, dynamic causal modeling, and phase slope index provide objective measures of neural interaction, they are often limited in their ability to capture the dynamic, task-aligned processes that characterize creativity. To address this, we propose a novel hybrid framework that combines the interpretability of traditional signal processing with the expressive power of deep learning. The proposed Generative Creativity Network (GCN) models temporal and spatial EEG features associated with creative cognition, while the Adaptive Creativity Enhancement Strategy (ACES) facilitates real-time feedback and optimization during creative tasks. This framework simulates the iterative process of idea generation, evaluation, and refinement by leveraging EEG-derived features linked to creativity dimensions such as originality, flexibility, and task relevance. Empirical validation across four diverse EEG/fMRI datasets demonstrates the model's effectiveness in capturing creativity-related neural dynamics and supporting adaptive interventions. Rather than inferring neural mechanisms or causal pathways directly, the model offers a computationally grounded approach to studying creative cognition through scalable and interpretable neural modeling.