测试大模型能否像人一样被文字引发时间错觉，发现它们表现反常但有依据。
The Machine's Internal Clock: Do LLMs Share Human Temporal Illusions?
- 用6684组叙事对测试五类时间错觉，看模型是否受文本暗示影响。
- 14个大模型在四类错觉中选对答案，远超人类（仅两类正确）。
- 模型靠检索文献而非真实感知，可能误把知识当直觉。
人类对时间的感知是主观的。已知的时间错觉表明，大脑依赖上下文和关系线索判断持续时间，而非直接追踪时间流逝。以往研究多基于视觉或听觉刺激。现有大模型的时间感知评估集中于事件持续时间估计或多步时序推理。本文构建了一个包含6,684对叙事的新基准，覆盖五类时间错觉，探究仅凭文字是否能诱发人类时间错觉。60名真人读者在五类错觉中仅在两类（文本提示明显时）选择预期场景。14个大模型在同一任务中却在四类错觉中做出符合文献预测的选择，与人类行为显著不同。推理轨迹分析显示约70%的回答明确引用心理学研究，表明这种一致性源于对已有文献的检索，而非类似人类的时空直觉。
原文摘要 · Abstract (English)
Human perception of time is subjective. Well-documented temporal illusions show that the brain relies on context and relational cues for judging duration instead of tracking elapsed time directly. Prior studies established these effects with visual and auditory stimuli. Existing LLM evaluations of temporal perception focus on estimating event durations or multi-step temporal reasoning. In this work, we investigate whether written narratives alone can evoke human temporal illusions, using a new benchmark of 6,684 narrative pairs spanning five illusions. We find that human readers (60 participants) prefer expected scenarios in only two of the five illusions, those where the manipulation is directly visible in text rather than requiring readers to internally simulate duration. We evaluate 14 LLMs on the same benchmark. Surprisingly, we find that models pick the literature-predicted scenario across four of the five illusions, diverging from human behavior. Reasoning traces show that ~70% of responses explicitly evoke psychology research, suggesting that this alignment is consistent with retrieval of published findings rather than human-like temporal biases.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。