On the Semantics of Large Language Models
This addresses a foundational controversy in AI about the nature of language understanding in LLMs, which is incremental as it builds on existing theories.
The paper tackles the problem of whether large language models (LLMs) truly understand language by examining their semantics at the word and sentence level, drawing on classical semantic theories to provide a nuanced assessment of their capabilities.
Large Language Models (LLMs) such as ChatGPT demonstrated the potential to replicate human language abilities through technology, ranging from text generation to engaging in conversations. However, it remains controversial to what extent these systems truly understand language. We examine this issue by narrowing the question down to the semantics of LLMs at the word and sentence level. By examining the inner workings of LLMs and their generated representation of language and by drawing on classical semantic theories by Frege and Russell, we get a more nuanced picture of the potential semantic capabilities of LLMs.