RKEFino1：一款法規知識增強型大型語言模型

摘要

近期大型語言模型（LLMs）的進展為金融應用帶來了巨大潛力，但同時在數字監管報告（DRR）領域引入了關鍵的準確性和合規性挑戰。為解決這些問題，我們提出了RKEFino1，這是一個基於Fino1構建的、融合了XBRL、CDM和MOF領域知識的監管知識增強型金融推理模型。我們設計了兩類問答任務——基於知識的推理和數學推理，並引入了一種新穎的數值命名實體識別（NER）任務，涵蓋了句子和表格中的金融實體。實驗結果證明了RKEFino1在合規性關鍵金融任務中的有效性和泛化能力。我們已將模型發佈於Hugging Face平台。

English

Recent advances in large language models (LLMs) hold great promise for financial applications but introduce critical accuracy and compliance challenges in Digital Regulatory Reporting (DRR). To address these issues, we propose RKEFino1, a regulation knowledge-enhanced financial reasoning model built upon Fino1, fine-tuned with domain knowledge from XBRL, CDM, and MOF. We formulate two QA tasks-knowledge-based and mathematical reasoning-and introduce a novel Numerical NER task covering financial entities in both sentences and tables. Experimental results demonstrate the effectiveness and generalization capacity of RKEFino1 in compliance-critical financial tasks. We have released our model on Hugging Face.

RKEFino1：一款法規知識增強型大型語言模型

RKEFino1: A Regulation Knowledge-Enhanced Large Language Model

摘要

Support