Scaling and context steer LLMs along the same computational path as the human brain

TL;DR

This study demonstrates that large language models and the human brain exhibit similar sequential processing patterns, with alignment influenced by model size and context length, revealing insights into their computational similarities.

Contribution

It provides empirical evidence that LLMs and the human brain process information in a similar layered sequence, influenced by model size and context, advancing understanding of neural and artificial computation alignment.

Findings

LLMs and brain responses align in a layer-wise sequential manner

Alignment depends on model size and context length

Both transformers and recurrent architectures show similar patterns

Abstract

Recent studies suggest that the representations learned by large language models (LLMs) are partially aligned to those of the human brain. However, whether and why this alignment score arises from a similar sequence of computations remains elusive. In this study, we explore this question by examining temporally-resolved brain signals of participants listening to 10 hours of an audiobook. We study these neural dynamics jointly with a benchmark encompassing 22 LLMs varying in size and architecture type. Our analyses confirm that LLMs and the brain generate representations in a similar order: specifically, activations in the initial layers of LLMs tend to best align with early brain responses, while the deeper layers of LLMs tend to best align with later brain responses. This brain-LLM alignment is consistent across transformers and recurrent architectures. However, its emergence depends on both model size and context length. Overall, this study sheds light on the sequential nature of computations and the factors underlying the partial convergence between biological and artificial neural networks.