Meaning without reference in large language models
This addresses philosophical skepticism about meaning in AI for researchers and theorists, but it is incremental as it builds on existing cognitive theories.
The paper argues that large language models (LLMs) likely capture aspects of meaning by approximating conceptual role theories from human cognition, suggesting this explains their success and potential for human-like improvement.
The widespread success of large language models (LLMs) has been met with skepticism that they possess anything like human concepts or meanings. Contrary to claims that LLMs possess no meaning whatsoever, we argue that they likely capture important aspects of meaning, and moreover work in a way that approximates a compelling account of human cognition in which meaning arises from conceptual role. Because conceptual role is defined by the relationships between internal representational states, meaning cannot be determined from a model's architecture, training data, or objective function, but only by examination of how its internal states relate to each other. This approach may clarify why and how LLMs are so successful and suggest how they can be made more human-like.