Piantadosi and Hill propose that reference is not a necessary condition for meaningful output in Large Language Models, because these models can operate within coherent systems of conceptual roles derived from their training data.

Authors

• Person: Not available Organization: arXiv
  Not Minds, but Signs: Reframing LLMs through Semiotics - arXiv

Sources

• Not Minds, but Signs: Reframing LLMs through Semiotics - arXiv arxiv.org arXiv via serper

Referenced by nodes (1)

• Large Language Models concept