In what sense does a large language model (LLM) have knowledge? We answer by granting LLMs ‘instrumental knowledge’: knowledge gained by using next-word generation as an instrument. We then ask how instrumental knowledge is related to the ordinary, ‘worldly knowledge’ exhibited by humans, and explore this question in terms of the degree to which instrumental knowledge can be said to incorporate the structured world models of cognitive science. We discuss ways LLMs could recover degrees of worldly knowledge and suggest that such recovery will be governed by an implicit, resource-rational tradeoff between world models and tasks. Our answer to this question extends beyond the capabilities of a particular AI system and challenges assumptions about the nature of knowledge and intelligence.

中文翻译：

---

从任务结构到世界模型：做什么LLMs知道？


大型语言模型（LLM ）有知识吗？我们通过授予来回答LLMs“工具知识”：通过使用下一代单词生成作为工具获得的知识。然后，我们询问工具性知识与人类所展示的普通“世俗知识”有何关系，并从工具性知识可以在多大程度上纳入认知科学的结构化世界模型的角度探讨这个问题。我们讨论方法LLMs可以恢复一定程度的世俗知识，并表明这种恢复将受到世界模型和任务之间隐含的、资源理性的权衡的控制。我们对这个问题的回答超出了特定人工智能系统的能力，并挑战了关于知识和智能本质的假设。