让摘要可追溯来源，避免AI胡编乱造
sui-1: Grounded and Verifiable Long-Form Summarization
- 生成带引用的摘要，每条结论都能对应原文句子
- 在5种语言上训练，超2.2万条高质量数据
- 小模型胜过大模型，适合法律政府等严谨场景
大语言模型常生成看似合理却不可验证的摘要，这对政府、法律等合规敏感领域构成重大风险。我们提出 sui-1，一个240亿参数的模型，可生成带有内联引用的摘要，使用户能将每个主张追溯至原始句子。其合成数据流水线结合思维链提示与多阶段验证，从议会文件、网页文本和维基百科等多种来源，生成了超过22,000条跨五种语言的高质量训练样本。评估显示，sui-1显著优于所有测试的开源模型，包括参数量达其三倍的模型。结果表明，针对任务的专项训练远胜于单纯扩大模型规模。模型权重与交互式演示已公开。
原文摘要 · Abstract (English)
Large language models frequently generate plausible but unfaithful summaries that users cannot verify against source text, a critical limitation in compliance-sensitive domains such as government and legal analysis. We present sui-1, a 24B parameter model that produces abstractive summaries with inline citations, enabling users to trace each claim to its source sentence. Our synthetic data pipeline combines chain-of-thought prompting with multi-stage verification, generating over 22,000 high-quality training examples across five languages from diverse sources including parliamentary documents, web text, and Wikipedia. Evaluation shows sui-1 significantly outperforms all tested open-weight baselines, including models with 3x more parameters. These results demonstrate that task-specific training substantially outperforms scale alone for citation-grounded summarization. Model weights and an interactive demo are publicly available.
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。