On the Other Side of the Mirror: How Language Models Align with the Human Brain | by Salvatore Raieli

| LLM | Neuro Science | Brain | Align |

Looking for linguistic qualifications and the evolution of brain alignment in large language models

This study analyzes the alignment of a large language model (LLM) with the human brain language network in 34 training checkpoints. This shows that the alignment of the brain closes closely to the linguistic linguistic ability, such as linguistic rules, while practical ability, including reasoning, develops less firmly. Research also shows that the size of the model is not a reliable predecessor of mental alignment and that the Future of better language performance highlights opportunities to improve future LLM. — The image created by the author using AI

If you speak to a man in a language he understands, he goes to his head. If you speak to her in her language, it goes to her heart. – Nelson Mandela

Understanding how the brain works and how the language acts are included in the basic goals of neuro science. Human language processing is supported by the human brain Network of Language (LN)A set of left -handed frontotimpurial regions in the brain. LN gives strong and selectively responds to the input of the language, and researchers have also tried to study using it Large language models (llms). The LLM is trained to predict the next token and appear to capture some aspects of the human reaction about the language.

Post Views: 9