FISCAL: 金融合成声明-文档增强学习用于高效事实核查
FISCAL: Financial Synthetic Claim-document Augmented Learning for Efficient Fact-Checking
AI总结：
FISCAL通过生成领域特定合成数据和高效微调，使小型模型在金融事实核查任务中达到最先进的准确性、鲁棒性和可扩展性。
摘要：
大型语言模型（LLMs）在金融领域的应用需要事实可靠性和计算效率，但当前系统常常产生细节幻觉并依赖于极大规模的模型。我们提出了FISCAL（金融合成声明-文档增强学习），一种生成针对金融事实核查的合成数据的模块化框架。使用FISCAL，我们生成了一个名为FISCAL-data的数据集，并用它来训练MiniCheck-FISCAL，一种轻量级的验证器用于数值金融声明。MiniCheck-FISCAL在基准测试中优于其基线，超越了GPT-3.5 Turbo和其他类似大小的开源系统，并接近更大系统的准确性（20倍），如Mixtral-8x22B和Command R+。在外部数据集FinDVer和Fin-Fact上，它与GPT-4o和Claude-3.5相当，但优于Gemini-1.5 Flash。这些结果表明，结合高效微调的领域特定合成数据使小型模型能够实现最先进的准确性、鲁棒性和可扩展性，以满足实际金融AI的需求。数据集和脚本可在项目存储库中获得（论文中提供链接）。
英文摘要：
Financial applications of large language models (LLMs) require factual reliability and computational efficiency, yet current systems often hallucinate details and depend on prohibitively large models. We propose FISCAL (Financial Synthetic Claim-Document Augmented Learning), a modular framework for generating synthetic data tailored to financial fact-checking. Using FISCAL, we generate a dataset called FISCAL-data and use it to train MiniCheck-FISCAL, a lightweight verifier for numerical financial claims. MiniCheck-FISCAL outperforms its baseline, surpasses GPT-3.5 Turbo and other open-source peers of similar size, and approaches the accuracy of much larger systems (20x), such as Mixtral-8x22B and Command R+. On external datasets FinDVer and Fin-Fact, it rivals GPT-4o and Claude-3.5 while outperforming Gemini-1.5 Flash. These results show that domain-specific synthetic data, combined with efficient fine-tuning, enables compact models to achieve state-of-the-art accuracy, robustness, and scalability for practical financial AI. The dataset and scripts are available in the project repository (link provided in the paper).
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。
