Tag
This paper explores if large language models exhibit human-like temporal illusions by testing them on a narrative benchmark, finding that LLMs often choose scenarios consistent with psychology literature rather than mimicking human perception.