LLM-inspired study sheds light on how brain cells encode the meaning of words
Understanding how the human brain processes and represents spoken language remains an important goal for neuroscientists. Understanding spoken words involves more than simply recognizing sounds—it also requires the brain to decipher the meaning of words and how those meanings change with context.
This research project, conducted by a team comprising scientists from Baylor College of Medicine, Rice University, the University of California, Berkeley, and Texas Children's Hospital, aimed to explore how the human brain encodes the meaning of words. Their study, published in Nature Neuroscience in 2026, utilized artificial intelligence systems inspired by large language models (LLMs) such as ChatGPT and Gemini to gain insights into the process.
The researchers recorded the activity of individual neurons in the hippocampus of ten individuals while they listened to narratives. They then used mathematical models to analyze the linguistic content of the stories and the recorded brain activity. The objective was to determine if information about word meaning could predict the patterns of neuronal activity.
The findings indicated that the meaning of words is encoded through patterns of activity across multiple neurons in the hippocampus. Individual words displayed complex preferences for certain meanings, suggesting a nuanced encoding process. The researchers compared the neural activity patterns with those generated by language models and found that the brain's response closely aligned with embeddings from a specific LLM, GPT-2.
Furthermore, the variability in neural responses correlated with the polysemy of words, indicating that semantic encoding is influenced by context.
In essence, this study provides evidence that the hippocampus encodes word meaning in a distributed manner, with the collective activity of groups of cells carrying information about word relationships and contextual variations. This discovery not only enhances our understanding of how the hippocampus processes language but also opens avenues for future research that could leverage LLMs to further investigate the complexities of human cognition.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
