无需微调，实现大模型零样本摘要长度精准控制
Zero-Shot Strategies for Length-Controllable Summarization
- 提出长度逼近、目标调整等四类零样本控制方法
- 在LLaMA 3上显著提升摘要长度合规率，质量不降反升
- 适合需要稳定输出长度的实用化摘要系统
大型语言模型在零样本设置下难以实现精确的长度控制。我们通过多维度评估模型的长度控制能力，发现不同度量方式下存在显著差异，并揭示了模型内在偏差。针对该问题，我们提出长度逼近、目标调整、样本筛选和自动修正四类方法。结合使用后，在不进行模型微调或结构修改的前提下，显著提升了摘要长度合规性，同时保持甚至提高了摘要质量。本研究不仅深化了对大模型可控生成行为的理解，也为真实场景中更可靠、可适应的摘要系统提供了有效方案。
原文摘要 · Abstract (English)
Large language models (LLMs) struggle with precise length control, particularly in zero-shot settings. We conduct a comprehensive study evaluating LLMs' length control capabilities across multiple measures and propose practical methods to improve controllability. Our experiments with LLaMA 3 reveal stark differences in length adherence across measures and highlight inherent biases of the model. To address these challenges, we introduce a set of methods: length approximation, target adjustment, sample filtering, and automated revisions. By combining these methods, we demonstrate substantial improvements in length compliance while maintaining or enhancing summary quality, providing highly effective zero-shot strategies for precise length control without the need for model fine-tuning or architectural changes. With our work, we not only advance our understanding of LLM behavior in controlled text generation but also pave the way for more reliable and adaptable summarization systems in real-world applications.
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。