A new study by Brown researchers provides evidence that the mechanisms of reasoning between humans and LLMs are similar.