无需训练的提示压缩框架，提升长文本中关键信息的识别效率。
Perception Compressor: A Training-Free Prompt Compression Framework in Long Context Scenarios
- 用引导问题和指令筛选最相关示例，精准定位关键信息。
- 动态分配压缩比例，在保留核心内容的同时减少冗余。
- 适用于长上下文场景，尤其适合对精度要求高的问答任务。
大型语言模型在多种场景下表现出色，但在长上下文任务中常因信息冗余且对关键信息位置敏感而表现下降。为此，我们提出 Perception Compressor——一种无需训练的提示压缩框架。该框架包含：基于引导问题与指令的感知检索器，用于提取最相关演示；双斜率比例分配器，动态调节压缩比与开书比；以及半引导迭代压缩机制，在保留关键标记的同时移除干扰项。我们在 NaturalQuestions、LongBench 与 MuSiQue 等长上下文基准上进行了大量实验，结果表明，Perception Compressor 显著优于现有方法，达到当前最佳性能。
原文摘要 · Abstract (English)
Large language models (LLMs) demonstrate exceptional capabilities in various scenarios. However, they suffer from much redundant information and are sensitive to the position of key information in long context scenarios. To address these challenges, we present Perception Compressor, a training-free prompt compression framework. It includes a perception retriever that leverages guiding questions and instruction to retrieve the most relevant demonstrations, a dual-slope ratio allocator to dynamically allocate compression ratios and open-book ratios, and a semi-guided iterative compression that retains key information at the token level while removing tokens that distract the LLM. We conduct extensive experiments on long context benchmarks, i.e., NaturalQuestions, LongBench, and MuSiQue. Experiment results show that Perception Compressor outperforms existing methods by a large margin, achieving state-of-the-art performance.
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。