Do Language Models Have Mental States

3 credits · Graduate

Do large language models (LLMs) understand what they say? Do they have desires? Do they feel pain or emotions? How can we know? In this class we will survey philisophical background on the nature of mental states and read recent papers on LLM representations. We will try to gain clarity about the nature of mental states and whether LLMs have them now or might in the future. We may stray into hard problems of consciousness and qualia.