Mapping Language in the Brain with AI: A Study on Semantic and Syntactic Representation in Neural Models and Human Brain Activity

Understanding how the human brain processes language has been a long-standing challenge in neuroscience and cognitive science. Recent advancements in artificial intelligence (AI), particularly in neural networks, have opened new avenues for investigating the representation of language in the brain. This study explores the relationship between semantic and syntactic representations in neural models and human brain activity. By comparing how deep learning models and the human brain process linguistic structures, this research seeks to bridge the gap between computational models and biological systems. The research aims to analyze the similarities and differences in how neural models and the human brain represent syntactic and semantic information. Using functional magnetic resonance imaging (fMRI) data from brain activity during language processing tasks, and applying AI models trained on large language datasets, this study investigates the neural correlates of syntax and semantics. The results show that certain regions of the brain correspond to the syntactic structures processed by AI models, while others align more closely with semantic representations. The neural network models exhibited high correspondence with brain activity patterns, particularly in tasks involving sentence structure and meaning comprehension. This study concludes that AI models can be used to enhance our understanding of how language is represented in the brain, offering valuable insights into both neuroscience and artificial intelligence.