大型语言模型如何被设计为产生幻觉
How Large Language Models are Designed to Hallucinate
AI总结：
本文提出LLM幻觉源于transformer架构的结构性限制而非数据缺陷，区分本体论幻觉与残余推理幻觉，并通过海德格尔框架和跨十二个模型的实验加以验证，最终提出真值约束架构的设计方向。
摘要：
大型语言模型（LLMs）在语言和推理任务中表现出卓越的流畅性，但仍然系统性地容易产生幻觉。主流观点将幻觉归因于数据缺口、有限的上下文或优化错误。我们则认为，幻觉是transformer架构的结构性结果。作为连贯性引擎，transformer被迫产生流畅的续写，其自注意力机制模拟了意义的关联结构，但缺乏时间性、情绪和关切的存在性根基，而这些正是稳定人类理解的基础。在此基础上，我们区分了本体论幻觉和残余推理幻觉：前者产生于续写需要揭示世界中的存在者时，后者则是模型通过回收文本中人类推理的痕迹来模仿推理。我们通过符合海德格尔范畴的案例研究，以及一项跨十二个LLMs的实验来说明这些模式，该实验展示了在扩展提示下模拟的“自我保存”如何出现。我们的贡献有三方面：（1）一个比较性说明，展示为何现有解释不充分；（2）一个与存在结构相关联的幻觉预测性分类法，并提出相应的基准；（3）面向“真值约束”架构的设计方向，这种架构能够在缺乏揭示时进行保留或延迟。我们得出结论：幻觉并非偶然缺陷，而是基于transformer的模型的一个决定性限制，是脚手架可以掩盖但永远无法解决的结果。
英文摘要：
Large language models (LLMs) achieve remarkable fluency across linguistic and reasoning tasks but remain systematically prone to hallucination. Prevailing accounts attribute hallucinations to data gaps, limited context, or optimization errors. We argue instead that hallucination is a structural outcome of the transformer architecture. As coherence engines, transformers are compelled to produce fluent continuations, with self-attention simulating the relational structure of meaning but lacking the existential grounding of temporality, mood, and care that stabilizes human understanding. On this basis, we distinguish ontological hallucination, arising when continuations require disclosure of beings in world, and residual reasoning hallucination, where models mimic inference by recycling traces of human reasoning in text. We illustrate these patterns through case studies aligned with Heideggerian categories and an experiment across twelve LLMs showing how simulated "self-preservation" emerges under extended prompts. Our contribution is threefold: (1) a comparative account showing why existing explanations are insufficient; (2) a predictive taxonomy of hallucination linked to existential structures with proposed benchmarks; and (3) design directions toward "truth-constrained" architectures capable of withholding or deferring when disclosure is absent. We conclude that hallucination is not an incidental defect but a defining limit of transformer-based models, an outcome scaffolding can mask but never resolve.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
