注意力溢出：长上下文缺失项推荐中的语言模型输入模糊
Attention Overflow: Language Model Input Blur during Long-Context Missing Items Recommendation
AI总结：
该研究揭示大型语言模型在长上下文缺失项推荐中因输入项目过多而产生“注意力溢出”现象，导致重复建议，并提出迭代循环可缓解但成本随重复率增加。
摘要：
大型语言模型（LLMs）能够从提示中列出的项目中建议缺失的元素，这可用于列表补全或基于用户历史的推荐。然而，当呈现过多项目时，其性能会下降，因为它们开始建议输入列表中已包含的项目。对于2024年年中的旗舰LLMs，这种情况发生在大约100个项目时。我们在合成问题（例如，在给定范围的乱序整数中查找缺失数字）和现实电影推荐场景中评估了这一现象。我们将此问题称为“注意力溢出”，因为防止重复需要同时关注所有项目。尽管迭代循环可以缓解此问题，但其成本随重复率增加而上升，影响语言模型从长输入中获取新颖性的能力。
英文摘要：
Large language models (LLMs) can suggest missing elements from items listed in a prompt, which can be used for list completion or recommendations based on users' history. However, their performance degrades when presented with too many items, as they start to suggest items already included in the input list. This occurs at around 100 items for mid-2024 flagship LLMs. We evaluate this phenomenon on both synthetic problems (e.g., finding missing numbers in a given range of shuffled integers) and realistic movie recommendation scenarios. We refer to this issue as \textit{attention overflow}, as preventing repetition requires attending to all items simultaneously. Although iterative loops can mitigate this problem, their costs increase with the repetition rate, affecting the language models' ability to derive novelty from lengthy inputs.
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。
