Conceptual structure coheres in human cognition but not in large language models

TL;DR

This study compares the conceptual structures of humans and large language models using cognitive psychology methods, revealing that LLMs show less consistent and more task-dependent structures than humans, highlighting fundamental differences.

Contribution

It introduces a novel comparative analysis of conceptual structures in humans and LLMs using cognitive psychology techniques, revealing key differences in structure stability and task dependence.

Findings

Human conceptual structures are robust across cultures and methods.

LLMs' conceptual structures vary significantly with the task used.

Humans show more consistent conceptual structures than LLMs.

Abstract

Neural network models of language have long been used as a tool for developing hypotheses about conceptual representation in the mind and brain. For many years, such use involved extracting vector-space representations of words and using distances among these to predict or understand human behavior in various semantic tasks. Contemporary large language models (LLMs), however, make it possible to interrogate the latent structure of conceptual representations using experimental methods nearly identical to those commonly used with human participants. The current work utilizes three common techniques borrowed from cognitive psychology to estimate and compare the structure of concepts in humans and a suite of LLMs. In humans, we show that conceptual structure is robust to differences in culture, language, and method of estimation. Structures estimated from LLM behavior, while individually fairly consistent with those estimated from human behavior, vary much more depending upon the particular task used to generate responses--across tasks, estimates of conceptual structure from the very same model cohere less with one another than do human structure estimates. These results highlight an important difference between contemporary LLMs and human cognition, with implications for understanding some fundamental limitations of contemporary machine language.