当大模型检索自己生成的内容时，回答会逐渐崩溃，失去多样性。
RAG Collapse: LLM Responses Collapse When Retrieved Documents Are Self-Authored
- 用自生成文档做检索，触发大模型回答崩溃
- 超79%实验中，系统因自引用导致回答失效
- 哪怕只有一条自生成参考，也会引发偏差
大模型的回答依赖互联网数据（通过训练或RAG），而如今大量网络内容由AI生成（Paredes et al., 2026），可能形成自我强化的反馈循环。已有研究发现，当大模型递归训练自身输出时，会出现模型崩溃（Shumailov et al., 2024）：回答多样性下降，最终偏离原始训练数据。本文揭示，若基于大模型的AI系统使用搜索工具检索其自身生成的参考文献，同样会发生类似崩溃，称之为RAG崩溃。我们对三种模拟场景、三类模型家族、1,019个信息查询提示进行了广泛实验，共完成1,528次模拟和超过一百万次LLM API调用，发现79.6%（1,216/1,528）的模拟出现崩溃。令人意外的是，仅一条自生成参考即可触发崩溃，因为大模型过度引用自身内容。这种自偏倚现象在控制参考质量后依然存在。
原文摘要 · Abstract (English)
LLM responses are based on the internet (via training or RAG), and AI is now used to generate a significant amount of content online (Paredes et al., 2026), creating the potential for a self-reinforcing feedback loop. Prior work has shown that when LLMs are recursively trained on their own output, they experience model collapse (Shumailov et al., 2024): responses become less diverse, and eventually no longer resemble the original training data. In this paper, we show that a similar collapse occurs if LLM-based AI systems retrieve references they authored using a search tool. We call this RAG collapse. We conduct extensive experiments with three types of simulations of AI systems retrieving references they generated, using three model families, and 1,019 information-seeking prompts, totaling 1,528 simulations and over one million LLM API calls, and find that 79.6% (1,216/1,528) of simulations end in collapse. Surprisingly, even a single self-authored reference can trigger collapse because the LLM disproportionately cites its own content. This self-bias persists even after controlling for reference quality.
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。