用词频前瞻提升大模型摘要忠实度，无需训练即可改进关键信息保留。
BLooP: Zero-Shot Abstractive Summarization using Large Language Models with Bigram Lookahead Promotion
- 通过词对前瞻机制，在生成时优先选择原文中高频共现的词组。
- 在CNN/DM等数据集上，ROUGE和BARTScore均显著提升，最高增益达6.3点。
- 无需训练或修改模型，适合希望快速提升摘要质量的研究者与应用开发者。
抽象摘要要求模型生成能准确传达源文档信息的摘要。尽管大语言模型可在无微调情况下生成摘要，但常遗漏关键细节并引入无关内容。我们提出BLooP（Bigram Lookahead Promotion），一种无需训练的解码干预方法，通过在每一步解码时进行词对哈希表查找，引导大语言模型生成源文档中出现过的词组。该方法不需训练、微调或修改模型结构。我们在Llama-3.1-8B-Instruct、Mistral-Nemo-Instruct-2407和Gemma-2-9b-it上测试了CNN/DM、CCSum、Multi-News和SciTLDR数据集，结果显示ROUGE和BARTScore均有提升。人工评估表明，BLooP显著提高了摘要忠实度，同时保持可读性。代码已开源：https://github.com/varuniyer/BLooP。
原文摘要 · Abstract (English)
Abstractive summarization requires models to generate summaries that convey information in the source document. While large language models can generate summaries without fine-tuning, they often miss key details and include extraneous information. We propose BLooP (Bigram Lookahead Promotion), a simple training-free decoding intervention that encourages large language models (LLMs) to generate tokens that form bigrams from the source document. BLooP operates through a hash table lookup at each decoding step, requiring no training, fine-tuning, or model modification. We demonstrate improvements in ROUGE and BARTScore for Llama-3.1-8B-Instruct, Mistral-Nemo-Instruct-2407, and Gemma-2-9b-it on CNN/DM, CCSum, Multi-News, and SciTLDR. Human evaluation shows that BLooP significantly improves faithfulness without reducing readability. We make the code available at https://github.com/varuniyer/BLooP
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。